multitask_intent.ipynb

{"cells":[{"cell_type":"code","execution_count":2,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":24320,"status":"ok","timestamp":1684514947930,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"WHwHkm4rGCLK","outputId":"82635074-961e-4964-abe0-fdffca206aec"},"outputs":[{"output_type":"stream","name":"stdout","text":["Mounted at /content/drive\n"]}],"source":["from google.colab import drive\n","drive.mount('/content/drive')"]},{"cell_type":"code","execution_count":3,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":7,"status":"ok","timestamp":1684514947930,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"VFg6mNX7Gdxg","outputId":"8caaeb47-a222-48e3-d3aa-4811807e2270"},"outputs":[{"output_type":"stream","name":"stdout","text":["Fri May 19 16:49:06 2023       \n","+-----------------------------------------------------------------------------+\n","| NVIDIA-SMI 525.85.12    Driver Version: 525.85.12    CUDA Version: 12.0     |\n","|-------------------------------+----------------------+----------------------+\n","| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n","| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n","|                               |                      |               MIG M. |\n","|===============================+======================+======================|\n","|   0  NVIDIA A100-SXM...  Off  | 00000000:00:04.0 Off |                    0 |\n","| N/A   30C    P0    45W / 400W |      0MiB / 40960MiB |      0%      Default |\n","|                               |                      |             Disabled |\n","+-------------------------------+----------------------+----------------------+\n","                                                                               \n","+-----------------------------------------------------------------------------+\n","| Processes:                                                                  |\n","|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n","|        ID   ID                                                   Usage      |\n","|=============================================================================|\n","|  No running processes found                                                 |\n","+-----------------------------------------------------------------------------+\n"]}],"source":["gpu_info = !nvidia-smi\n","gpu_info = '\\n'.join(gpu_info)\n","if gpu_info.find('failed') >= 0:\n","  print('Not connected to a GPU')\n","else:\n","  print(gpu_info)"]},{"cell_type":"code","execution_count":4,"metadata":{"id":"ZQU-d0knGyhl","executionInfo":{"status":"ok","timestamp":1684514952855,"user_tz":-330,"elapsed":4927,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["import os \n","import numpy as np\n","import pandas as pd\n","import json\n","import warnings\n","import logging\n","import gc\n","import random\n","import math\n","import re\n","import ast\n","from tqdm import tqdm\n","from typing import Optional\n","from datetime import datetime\n","import pickle\n","import sklearn\n","from torch.utils.data import Dataset, DataLoader\n","from sklearn.metrics import jaccard_score, f1_score, accuracy_score, recall_score, precision_score, confusion_matrix\n","from sklearn.model_selection import train_test_split\n","import copy"]},{"cell_type":"code","execution_count":5,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":3706,"status":"ok","timestamp":1684514956560,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"Ul7wZ5-xCDe8","outputId":"4a05bc4c-be58-453c-fe3c-d8e9772b44fc"},"outputs":[{"output_type":"stream","name":"stdout","text":["Name: tqdm\n","Version: 4.65.0\n","Summary: Fast, Extensible Progress Meter\n","Home-page: https://tqdm.github.io\n","Author: \n","Author-email: \n","License: MPLv2.0, MIT Licences\n","Location: /usr/local/lib/python3.10/dist-packages\n","Requires: \n","Required-by: cmdstanpy, gdown, hyperopt, kaggle, moviepy, nltk, panel, proglog, prophet, spacy, tensorflow-datasets, torchtext\n"]}],"source":["pip show tqdm"]},{"cell_type":"code","execution_count":6,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":16,"status":"ok","timestamp":1684514956561,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"z7LivYCPCYqZ","outputId":"01d02c20-e380-4194-9eb7-1809e7941fe1"},"outputs":[{"output_type":"stream","name":"stdout","text":["4.0\n"]}],"source":["print(pickle.format_version)"]},{"cell_type":"code","execution_count":7,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":13,"status":"ok","timestamp":1684514956561,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"jECujxkkCrsu","outputId":"a045a677-aac8-4443-de73-31a72adca16e"},"outputs":[{"output_type":"stream","name":"stdout","text":["3.10.11\n"]}],"source":["import platform\n","print(platform.python_version())"]},{"cell_type":"code","execution_count":8,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":10,"status":"ok","timestamp":1684514956561,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"kWnU1VcDC-Z3","outputId":"667fd328-bd47-400b-8694-2dbdc235881e"},"outputs":[{"output_type":"stream","name":"stdout","text":["1.2.2\n"]}],"source":["print(sklearn.__version__)"]},{"cell_type":"code","execution_count":9,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":35},"executionInfo":{"elapsed":9,"status":"ok","timestamp":1684514956562,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"nkDZ1EwRDLVt","outputId":"edab8117-a9c1-4c51-cd56-2ff8d7e2c1a1"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["'1.22.4'"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"string"}},"metadata":{},"execution_count":9}],"source":["np.__version__"]},{"cell_type":"code","execution_count":10,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":10040,"status":"ok","timestamp":1684514966594,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"suRa-hBHG_21","outputId":"73df8e49-28fd-4ac5-ef89-70fcd42434b3"},"outputs":[{"output_type":"stream","name":"stdout","text":["Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Collecting transformers\n","  Downloading transformers-4.29.2-py3-none-any.whl (7.1 MB)\n","\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m7.1/7.1 MB\u001b[0m \u001b[31m58.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n","\u001b[?25hRequirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from transformers) (3.12.0)\n","Collecting huggingface-hub<1.0,>=0.14.1 (from transformers)\n","  Downloading huggingface_hub-0.14.1-py3-none-any.whl (224 kB)\n","\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m224.5/224.5 kB\u001b[0m \u001b[31m26.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n","\u001b[?25hRequirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.10/dist-packages (from transformers) (1.22.4)\n","Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from transformers) (23.1)\n","Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from transformers) (6.0)\n","Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.10/dist-packages (from transformers) (2022.10.31)\n","Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from transformers) (2.27.1)\n","Collecting tokenizers!=0.11.3,<0.14,>=0.11.1 (from transformers)\n","  Downloading tokenizers-0.13.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (7.8 MB)\n","\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m7.8/7.8 MB\u001b[0m \u001b[31m86.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n","\u001b[?25hRequirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.10/dist-packages (from transformers) (4.65.0)\n","Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.14.1->transformers) (2023.4.0)\n","Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.14.1->transformers) (4.5.0)\n","Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (1.26.15)\n","Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (2022.12.7)\n","Requirement already satisfied: charset-normalizer~=2.0.0 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (2.0.12)\n","Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (3.4)\n","Installing collected packages: tokenizers, huggingface-hub, transformers\n","Successfully installed huggingface-hub-0.14.1 tokenizers-0.13.3 transformers-4.29.2\n"]}],"source":["!pip install transformers"]},{"cell_type":"code","execution_count":11,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":4638,"status":"ok","timestamp":1684514971228,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"r_Ovlo42G4Kh","outputId":"858f16a2-137c-4cac-c88a-50aefb928404"},"outputs":[{"output_type":"stream","name":"stdout","text":["Using GPU\n"]}],"source":["import nltk\n","from nltk.translate.bleu_score import sentence_bleu\n","from nltk.translate.meteor_score import meteor_score\n","# from rouge_score.rouge_scorer import RougeScorer\n","\n","warnings.filterwarnings(\"ignore\")\n","\n","import torch\n","from torch.utils.data import DataLoader, TensorDataset\n","\n","from transformers import (\n","    BartTokenizerFast,\n","    AdamW\n",")\n","\n","if torch.cuda.is_available():\n","    DEVICE = torch.device(\"cuda\")\n","    print(\"Using GPU\")\n","\n","else:\n","    DEVICE = torch.device(\"cpu\")\n","    print(\"Using CPU\")\n","\n","\n","from transformers import BertTokenizer, BertModel\n","\n","SOURCE_MAX_LEN = 234\n","\n","\n","ACOUSTIC_DIM = 768\n","ACOUSTIC_MAX_LEN = 1000\n","\n","\n","\n","VISUAL_DIM = 2048\n","VISUAL_MAX_LEN = 480\n","\n","\n","\n","import random\n","\n","\n","\n","\n","LEARNING_RATE = 1e-4\n","\n","\n"]},{"cell_type":"code","execution_count":12,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":48,"status":"ok","timestamp":1684514971228,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"cMdRaTfAHF7Q","outputId":"bd566214-fb8d-40e6-c981-6f4b74e90b25"},"outputs":[{"output_type":"stream","name":"stdout","text":["Seed : 994\n"]}],"source":["def set_random_seed(seed: int):\n","    print(\"Seed : {}\".format(seed))\n","\n","    torch.backends.cudnn.benchmark = False\n","    torch.backends.cudnn.enabled = False\n","    torch.backends.cudnn.deterministic = True\n","\n","    random.seed(seed)\n","    os.environ['PYTHONHASHSEED'] = str(seed)\n","    np.random.seed(seed)\n","    torch.manual_seed(seed)\n","    torch.cuda.manual_seed(seed)\n","    torch.cuda.manual_seed_all(seed)\n","\n","# set_random_seed(42)\n","# a = np.random.randint(0, 1000)\n","\n","# set_random_seed(123)\n","set_random_seed(994)\n","# set_random_seed(12345)\n","\n","import torch.nn as nn\n","import torch.nn.functional as F\n","import torch.utils.checkpoint\n","from torch.nn import CrossEntropyLoss, MSELoss\n","\n","from typing import Any, Callable, Dict, Iterable, List, Optional, Tuple, Union\n","\n","from transformers.modeling_utils import PreTrainedModel, unwrap_model\n","\n","from transformers import (\n","    BartTokenizer,\n","    AdamW\n",")\n","import transformers\n","from transformers.models.bart.configuration_bart import BartConfig\n","\n","from transformers.models.bart.modeling_bart import (\n","    BartPretrainedModel,\n","    BartDecoder,\n","    BartLearnedPositionalEmbedding,\n","    BartEncoderLayer,\n","    shift_tokens_right,\n","    _make_causal_mask,\n","    _expand_mask\n",")\n","\n","from transformers.modeling_outputs import (\n","    BaseModelOutput,\n","    Seq2SeqLMOutput,\n","    Seq2SeqModelOutput,\n","    Seq2SeqSequenceClassifierOutput\n",")"]},{"cell_type":"code","execution_count":13,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":45,"status":"ok","timestamp":1684514971228,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"oRyHg4XSBJ9J","outputId":"701de09d-3140-4a1c-f712-26e0ee3ac45e"},"outputs":[{"output_type":"stream","name":"stdout","text":["4.29.2\n"]}],"source":["print(transformers.__version__)"]},{"cell_type":"code","execution_count":14,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":43,"status":"ok","timestamp":1684514971229,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"7LFoXwXrB5UE","outputId":"3e5505e6-824a-40e1-d88c-032e9b9a1de3"},"outputs":[{"output_type":"stream","name":"stdout","text":["2.0.1+cu118\n"]}],"source":["print(torch.__version__)"]},{"cell_type":"code","execution_count":15,"metadata":{"id":"MuNh3gFEHQhY","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":41,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# from transformer_encoder import TransformerEncoder"]},{"cell_type":"code","execution_count":16,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":41,"status":"ok","timestamp":1684514971229,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"-ML9b3x4axdj","outputId":"6c3ff2e1-1ca0-4a41-f565-77bf76145746"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["device(type='cuda')"]},"metadata":{},"execution_count":16}],"source":["device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n","device"]},{"cell_type":"code","execution_count":17,"metadata":{"id":"OtwuoBrcwzUN","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# bert_tokenizer  = BertTokenizer.from_pretrained(\"bert-base-cased\")\n","# bert_model = BertModel.from_pretrained(\"bert-base-cased\")\n","# # bert_model"]},{"cell_type":"code","execution_count":18,"metadata":{"id":"e1XR3BlgHOY8","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["BART_INPUTS_DOCSTRING = r\"\"\"\n","    Args:\n","        input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):\n","            Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide\n","            it.\n","            Indices can be obtained using [`BartTokenizer`]. See [`PreTrainedTokenizer.encode`] and\n","            [`PreTrainedTokenizer.__call__`] for details.\n","            [What are input IDs?](../glossary#input-ids)\n","        attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):\n","            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:\n","            - 1 for tokens that are **not masked**,\n","            - 0 for tokens that are **masked**.\n","            [What are attention masks?](../glossary#attention-mask)\n","        decoder_input_ids (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):\n","            Indices of decoder input sequence tokens in the vocabulary.\n","            Indices can be obtained using [`BartTokenizer`]. See [`PreTrainedTokenizer.encode`] and\n","            [`PreTrainedTokenizer.__call__`] for details.\n","            [What are decoder input IDs?](../glossary#decoder-input-ids)\n","            Bart uses the `eos_token_id` as the starting token for `decoder_input_ids` generation. If `past_key_values`\n","            is used, optionally only the last `decoder_input_ids` have to be input (see `past_key_values`).\n","            For translation and summarization training, `decoder_input_ids` should be provided. If no\n","            `decoder_input_ids` is provided, the model will create this tensor by shifting the `input_ids` to the right\n","            for denoising pre-training following the paper.\n","        decoder_attention_mask (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):\n","            Default behavior: generate a tensor that ignores pad tokens in `decoder_input_ids`. Causal mask will also\n","            be used by default.\n","            If you want to change padding behavior, you should read [`modeling_bart._prepare_decoder_attention_mask`]\n","            and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more\n","            information on the default strategy.\n","        head_mask (`torch.Tensor` of shape `(encoder_layers, encoder_attention_heads)`, *optional*):\n","            Mask to nullify selected heads of the attention modules in the encoder. Mask values selected in `[0, 1]`:\n","            - 1 indicates the head is **not masked**,\n","            - 0 indicates the head is **masked**.\n","        decoder_head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):\n","            Mask to nullify selected heads of the attention modules in the decoder. Mask values selected in `[0, 1]`:\n","            - 1 indicates the head is **not masked**,\n","            - 0 indicates the head is **masked**.\n","        cross_attn_head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):\n","            Mask to nullify selected heads of the cross-attention modules in the decoder. Mask values selected in `[0,\n","            1]`:\n","            - 1 indicates the head is **not masked**,\n","            - 0 indicates the head is **masked**.\n","        encoder_outputs (`tuple(tuple(torch.FloatTensor)`, *optional*):\n","            Tuple consists of (`last_hidden_state`, *optional*: `hidden_states`, *optional*: `attentions`)\n","            `last_hidden_state` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) is a sequence of\n","            hidden-states at the output of the last layer of the encoder. Used in the cross-attention of the decoder.\n","        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):\n","            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape\n","            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape\n","            `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.\n","            Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention\n","            blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.\n","            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that\n","            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all\n","            `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of shape\n","            `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids` you\n","            can choose to directly pass an embedded representation. This is useful if you want more control over how to\n","            convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.\n","        decoder_inputs_embeds (`torch.FloatTensor` of shape `(batch_size, target_sequence_length, hidden_size)`, *optional*):\n","            Optionally, instead of passing `decoder_input_ids` you can choose to directly pass an embedded\n","            representation. If `past_key_values` is used, optionally only the last `decoder_inputs_embeds` have to be\n","            input (see `past_key_values`). This is useful if you want more control over how to convert\n","            `decoder_input_ids` indices into associated vectors than the model's internal embedding lookup matrix.\n","            If `decoder_input_ids` and `decoder_inputs_embeds` are both unset, `decoder_inputs_embeds` takes the value\n","            of `inputs_embeds`.\n","        use_cache (`bool`, *optional*):\n","            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see\n","            `past_key_values`).\n","        output_attentions (`bool`, *optional*):\n","            Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned\n","            tensors for more detail.\n","        output_hidden_states (`bool`, *optional*):\n","            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for\n","            more detail.\n","        return_dict (`bool`, *optional*):\n","            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.\n","\"\"\"\n","\n"]},{"cell_type":"markdown","metadata":{"id":"Q6aH8X1qjYgx"},"source":["#Sarcasm Only"]},{"cell_type":"code","execution_count":18,"metadata":{"id":"IhSj1MuFjYCm","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":[]},{"cell_type":"code","execution_count":19,"metadata":{"id":"aEDR_7N6HWsu","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class ContextAwareAttention(nn.Module):\n","\n","    def __init__(self,\n","                 dim_model : int,\n","                 dim_context : int,\n","                 dropout_rate : Optional[float] = 0.0 ):\n","\n","        super(ContextAwareAttention, self).__init__()\n","\n","        self.dim_model = dim_model\n","        self.dim_context = dim_context\n","        self.dropout_rate = dropout_rate\n","        self.attention_layer = nn.MultiheadAttention(embed_dim=self.dim_model,\n","                                                     num_heads = 1,\n","                                                     dropout = self.dropout_rate,\n","                                                     bias = True,\n","                                                    add_zero_attn=False,\n","                                                    batch_first=True,\n","                                                    device=DEVICE   \n","        )\n","\n","        self.u_k = nn.Linear(self.dim_context, self.dim_model, bias = False)\n","        self.w1_k = nn.Linear(self.dim_model, 1, bias=False)\n","        self.w2_k = nn.Linear(self.dim_model, 1, bias=False)\n","\n","        self.u_v = nn.Linear(self.dim_context, self.dim_model, bias=False)\n","        self.w1_v = nn.Linear(self.dim_model, 1, bias = False)\n","        self.w2_v = nn.Linear(self.dim_model, 1, bias = False)\n","\n","    def forward(self, q, k, v, context):\n","\n","        # print(\"Context shape : \", context.shape)\n","        # print(\"Dim context : \", self.dim_context, \" : Dim model : \", self.dim_model)\n","        key_context = self.u_k(context)\n","        # print(\"Context shape below key context : \", key_context.shape)\n","        value_context = self.u_v(context)\n","\n","        lambda_k = F.sigmoid(self.w1_k(k) + self.w2_k(key_context))\n","        lambda_v = F.sigmoid(self.w1_v(v) + self.w2_v(value_context))\n","\n","        k_cap = (1-lambda_k) * k + (lambda_k) * key_context\n","        v_cap = (1-lambda_v) * v + (lambda_v) * value_context\n","\n","        attention_output, _ = self.attention_layer(query = q,\n","                                                   key = k_cap,\n","                                                   value = v_cap)\n","  \n","        return attention_output                                     \n","\n"]},{"cell_type":"code","execution_count":20,"metadata":{"id":"2ZPn4U3Weq9I","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MAF_acoustic(nn.Module):\n","    def __init__(self, \n","                dim_model,\n","                dropout_rate):\n","        super(MAF_acoustic, self).__init__()\n","        self.dropout_rate = dropout_rate\n","\n","        self.acoustic_context_transform = nn.Linear(ACOUSTIC_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","        # self.visual_context_transform = nn.Linear(VISUAL_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","\n","        self.acoustic_context_attention = ContextAwareAttention(dim_model=dim_model,\n","                                                                dim_context=ACOUSTIC_DIM,\n","                                                                dropout_rate=dropout_rate)\n","\n","        # self.visual_context_attention = ContextAwareAttention(dim_model=dim_model,\n","        #                                                     dim_context=VISUAL_DIM,\n","        #                                                     dropout_rate=dropout_rate)\n","\n","        self.acoustic_gate = nn.Linear(2*dim_model, dim_model)\n","        # self.visual_gate = nn.Linear(2*dim_model, dim_model)\n","        self.dropout_layer = nn.Dropout(dropout_rate)\n","        self.final_layer_norm = nn.LayerNorm(dim_model)\n","\n","    def forward(self,\n","                text_input,\n","                acoustic_context):\n","\n","        # print(\"Acoustic context shape (A) : \", acoustic_context.shape)        \n","\n","        acoustic_context = acoustic_context.permute(0,2,1)\n","        acoustic_context = self.acoustic_context_transform(acoustic_context.float())\n","        acoustic_context = acoustic_context.permute(0,2,1)\n","\n","        audio_out = self.acoustic_context_attention(q=text_input,\n","                                                    k=text_input,\n","                                                    v=text_input,\n","                                                    context=acoustic_context)\n","        # print(\"Audio out (A) : \", audio_out.shape) \n","\n","        # print(\"Visual context shape : \", visual_context.shape)\n","        # visual_context = visual_context.permute(0,2,1)\n","        # visual_context = self.visual_context_transform(visual_context.float())\n","        # visual_context = visual_context.permute(0,2,1)\n","        \n","        # video_out = self.visual_context_attention(q=text_input,\n","        #                                             k=text_input,\n","        #                                             v=text_input,\n","        #                                             context=visual_context)\n","\n","        # print(\"Video out shape : \", video_out.shape)\n","        # print(\"Text input shape : \", text_input.shape)\n","        weight_a = F.sigmoid(self.acoustic_gate(torch.cat([text_input, audio_out], dim=-1)))\n","        # weight_v = F.sigmoid(self.visual_gate(torch.cat([text_input, video_out], dim=-1)))\n","\n","        # output = self.final_layer_norm(text_input + weight_a * audio_out + weight_v * video_out)\n","\n","        output = self.final_layer_norm(text_input + weight_a * audio_out)\n","\n","        return output"]},{"cell_type":"code","execution_count":21,"metadata":{"id":"ZeZdIlcker1V","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MAF_visual(nn.Module):\n","    def __init__(self, \n","                dim_model,\n","                dropout_rate):\n","        super(MAF_visual, self).__init__()\n","        self.dropout_rate = dropout_rate\n","\n","        # self.acoustic_context_transform = nn.Linear(ACOUSTIC_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","        self.visual_context_transform = nn.Linear(VISUAL_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","\n","        # self.acoustic_context_attention = ContextAwareAttention(dim_model=dim_model,\n","        #                                                         dim_context=ACOUSTIC_DIM,\n","        #                                                         dropout_rate=dropout_rate)\n","\n","        self.visual_context_attention = ContextAwareAttention(dim_model=dim_model,\n","                                                            dim_context=VISUAL_DIM,\n","                                                            dropout_rate=dropout_rate)\n","\n","        # self.acoustic_gate = nn.Linear(2*dim_model, dim_model)\n","        self.visual_gate = nn.Linear(2*dim_model, dim_model)\n","        self.dropout_layer = nn.Dropout(dropout_rate)\n","        self.final_layer_norm = nn.LayerNorm(dim_model)\n","\n","    def forward(self,\n","                text_input,\n","                visual_context):\n","\n","        # print(\"Acoustic context shape (A) : \", acoustic_context.shape)        \n","\n","        # acoustic_context = acoustic_context.permute(0,2,1)\n","        # acoustic_context = self.acoustic_context_transform(acoustic_context.float())\n","        # acoustic_context = acoustic_context.permute(0,2,1)\n","\n","        # audio_out = self.acoustic_context_attention(q=text_input,\n","        #                                             k=text_input,\n","        #                                             v=text_input,\n","        #                                             context=acoustic_context)\n","        # print(\"Audio out (A) : \", audio_out.shape) \n","\n","        # print(\"Visual context shape : \", visual_context.shape)\n","        visual_context = visual_context.permute(0,2,1)\n","        visual_context = self.visual_context_transform(visual_context.float())\n","        visual_context = visual_context.permute(0,2,1)\n","        \n","        video_out = self.visual_context_attention(q=text_input,\n","                                                    k=text_input,\n","                                                    v=text_input,\n","                                                    context=visual_context)\n","\n","        # print(\"Video out shape : \", video_out.shape)\n","        # print(\"Text input shape : \", text_input.shape)\n","        # weight_a = F.sigmoid(self.acoustic_gate(torch.cat([text_input, audio_out], dim=-1)))\n","        weight_v = F.sigmoid(self.visual_gate(torch.cat([text_input, video_out], dim=-1)))\n","\n","        # output = self.final_layer_norm(text_input + weight_a * audio_out + weight_v * video_out)\n","\n","        output = self.final_layer_norm(text_input  + weight_v * video_out)\n","\n","        return output"]},{"cell_type":"code","source":[],"metadata":{"id":"lkFzM91VBYIh","executionInfo":{"status":"ok","timestamp":1684514971229,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":21,"outputs":[]},{"cell_type":"code","execution_count":22,"metadata":{"id":"IVX99ZEfRebx","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class CustomTransformerEncoder(torch.nn.Module):\n","#   def __init__(self, vocab_size, hidden_dim, expand_factor):\n","#     super(CustomTransformerEncoder, self).__init__()\n"]},{"cell_type":"code","execution_count":23,"metadata":{"id":"lmfR5TqF20m4","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class MultiHeadSelfAttention(torch.nn.Module):\n","#   def __init__(self, config):\n","#     super(MultiHeadSelfAttention, self).__init__()\n","\n","#     self.num_attention_heads = 12\n","#     self.attention_head_size = config.d_model//12\n","#     self.embed_dim = config.d_model\n","#     self.dropout = config.dropout\n","\n","#     self.query = nn.Linear(config.d_model, config.d_model)\n","#     self.key = nn.Linear(config.d_model, config.d_model)\n","#     self.value = nn.Linear(config.d_model, config.d_model)\n","#     self.out_proj = nn.Linear(config.d_model, config.d_model)\n","\n","#   def forward(self, hidden_states, attention_mask):\n","#     query_states = self.query(hidden_states)\n","#     key_states = self.key(hidden_states)\n","#     value_states = self.value(hidden_states)\n","\n","#     batch_size = hidden_states.shape[0]\n","#     seq_len = hidden_states.shape[1]\n","\n","#     query_states =  query_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","#     key_states =  key_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","#     value_states = value_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","\n","#     query_states = query_states.view(batch_size*self.num_attention_heads, -1, self.attention_head_size)\n","#     key_states  = key_states.view(batch_size*self.num_attention_heads, -1, self.attention_head_size)\n","#     value_states = value_states.view(batch_size*self.num_attention_heads, -1, self.attention_head_size)\n","\n","#     attn_weights = torch.bmm(query_states, key_states.transpose(1,2))\n","\n","#     attn_weights = attn_weights.view(batch_size, self.num_attention_heads, attn_weights.shape[1], attn_weights.shape[2])\n","#     attn_weights = attn_weights/torch.sqrt(torch.tensor(self.attention_head_size))\n","\n","#     attn_weights_masked = attn_weights.masked_fill(\n","#         attention_mask[:, None, None, :] == False, -1e16\n","#     )\n","\n","#     attn_weights = torch.nn.functional.softmax(attn_weights_masked, dim = -1)\n","\n","#     attn_weights = torch.nn.functional.dropout(attn_weights, p = self.dropout, training = self.training)\n","\n","#     attn_weights = attn_weights.view(batch_size*self.num_attention_heads, seq_len, seq_len)\n","\n","\n","    \n","#     attn_output = torch.bmm(attn_weights, value_states)\n","\n","#     attn_output = attn_output.view(batch_size, self.num_attention_heads, seq_len, self.attention_head_size)\n","#     attn_output = attn_output.transpose(1,2)\n","\n","#     attn_output = attn_output.reshape(batch_size, seq_len, self.embed_dim)\n","\n","#     attn_out = self.out_proj(attn_output)\n","\n","#     return attn_out"]},{"cell_type":"code","execution_count":24,"metadata":{"id":"y0mlADX3NWj-","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# def gelu(input):\n","#   input = torch.tensor(input)\n","#   return input * 0.5 * (1.0 + torch.erf(input/torch.sqrt(torch.tensor(2.0))))"]},{"cell_type":"code","execution_count":25,"metadata":{"id":"DCJvPt5hFS3X","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":39,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class CustomEncoder(torch.nn.Module):\n","#   def __init__(self, config):\n","#     super(CustomEncoder, self).__init__()\n","#     self.embed_dim = config.d_model\n","#     self.self_attn = MultiHeadSelfAttention(config)\n","#     self.self_attn_layer_norm = nn.LayerNorm(self.embed_dim)\n","#     self.dropout = config.dropout\n","#     # self.activation_fn = torch.nn.ReLU()\n","#     self.activation_dropout = config.activation_dropout\n","#     self.fc1 = nn.Linear(self.embed_dim, 2*self.embed_dim)\n","#     self.fc2 = nn.Linear(2*self.embed_dim, self.embed_dim)\n","#     self.final_layer_norm = nn.LayerNorm(self.embed_dim)\n","\n","#   def forward(self, hidden_states, attention_mask):\n","#    residual = hidden_states\n","#    hidden_states =  self.self_attn(hidden_states = hidden_states, attention_mask = attention_mask)\n","#    hidden_states = nn.functional.dropout(hidden_states, p = self.dropout, training = self.training)\n","#    hidden_states = residual + hidden_states\n","#    hidden_states = self.self_attn_layer_norm(hidden_states)\n","\n","#    residual = hidden_states\n","#    hidden_states = gelu(self.fc1(hidden_states))\n","#    hidden_states = nn.functional.dropout(hidden_states, p=self.activation_dropout, training = self.training)\n","#    hidden_states = self.fc2(hidden_states)\n","#    hidden_states = nn.functional.dropout(hidden_states, p=self.activation_dropout, training = self.training)\n","#    hidden_states = residual + hidden_states\n","#    hidden_states = self.final_layer_norm(hidden_states)\n","\n","#    if hidden_states.dtype == torch.float16 and (\n","#        torch.isinf(hidden_states).any() or torch.isnan(hidden_states).any()\n","#    ):\n","#       clamp_value = torch.finfo(hidden_states.dtype).max = -1000\n","#       hidden_states = torch.clamp(hidden_states, min = -clamp_value, max = clamp_value)\n","\n","#    return hidden_states   "]},{"cell_type":"code","execution_count":26,"metadata":{"id":"kEWqLUTHbj0t","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class ContextEncoder(nn.Module):\n","#   def __init__(self, config, num_layers):\n","#     super(ContextEncoder, self).__init__()\n","\n","#     self.word_embeddings = nn.Embedding(config.vocab_size, config.d_model, padding_idx = config.pad_token_id)\n","#     self.position_embedding = nn.Embedding(500, config.d_model)\n","#     self.layer_norm = nn.LayerNorm(config.d_model)\n","#     self.context_encoder = torch.nn.ModuleList([CustomEncoder(config) for _ in range(num_layers)])\n","#     self.context_gate = nn.Linear(2*config.d_model, config.d_model)\n","    \n","\n","#   def forward(self, hidden_states, context_input_ids, context_attention_mask) :\n","    \n","#     seq_len = context_input_ids.shape[1]\n","#     pos = torch.arange(seq_len, dtype = torch.long)\n","#     pos = pos.unsqueeze(0).expand_as(context_input_ids)\n","#     pos = pos.to(device)\n","#     context_state = self.word_embeddings(context_input_ids) + self.position_embedding(pos)\n","    \n","#     for layer in self.context_encoder:\n","#             context_state = layer(hidden_states = context_state, attention_mask  = context_attention_mask)\n","    \n","\n","#     weight_c = F.sigmoid(self.context_gate(torch.cat([hidden_states, context_state], dim=-1)))\n","#     output = self.layer_norm(hidden_states + weight_c * context_state)\n","#     # print('transformer context shape :', output.shape)\n","#     return output\n","\n"]},{"cell_type":"code","execution_count":27,"metadata":{"id":"7Oa75YDExmAz","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class ContextEncoder(nn.Module):\n","#   def __init__(self, config):\n","#     super(ContextEncoder, self).__init__()\n","\n","#     self.bert = bert_model\n","#     self.layer_norm = nn.LayerNorm(config.d_model)\n","#     self.context_gate = nn.Linear(2*config.d_model, config.d_model)\n","    \n","\n","#   def forward(self, hidden_states, context_input_ids, context_attention_mask) :\n","    \n","#     output_encoding = self.bert(context_input_ids, context_attention_mask)\n","\n","#     context_state = output_encoding['last_hidden_state']\n","    \n","\n","#     weight_c = F.sigmoid(self.context_gate(torch.cat([hidden_states, context_state], dim=-1)))\n","#     output = self.layer_norm(hidden_states + weight_c * context_state)\n","#     # print('transformer context shape :', output.shape)\n","#     return output\n","\n"]},{"cell_type":"code","execution_count":28,"metadata":{"id":"s88m19VoHbj5","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MultiModalBartEncoder(BartPretrainedModel):\n","\n","    def __init__(self, config: BartConfig, embed_tokens: Optional[nn.Embedding] = None):\n","        super().__init__(config)\n","\n","        self.dropout = config.dropout\n","        self.layerdrop = config.encoder_layerdrop\n","\n","        embed_dim = config.d_model\n","        self.padding_idx = config.pad_token_id\n","        self.max_source_position = config.max_position_embeddings\n","        self.embed_scale = math.sqrt(embed_dim) if config.scale_embedding else 1.0\n","\n","        if embed_tokens is not None:\n","            self.embed_tokens = embed_tokens\n","        else:\n","            self.embed_tokens = nn.Embedding(config.vocab_size, embed_dim, self.padding_idx)\n","\n","        self.embed_positions = BartLearnedPositionalEmbedding(\n","            config.max_position_embeddings,\n","            embed_dim\n","        )        \n","\n","        \n","        \n","\n","        self.layers = nn.ModuleList([BartEncoderLayer(config) for _ in range(config.encoder_layers)])\n","\n","        self.layernorm_embedding = nn.LayerNorm(embed_dim)\n","\n","        self.init_weights()\n","        self.gradient_checkpointing = False\n","\n","        # self.fusion_at_layer = [4]\n","        # self.fusion_at_layer = [3, 4]\n","        # self.fusion_at_layer3 = [3]\n","        self.fusion_at_layer4 = [4]\n","        self.fusion_at_layer5 = [5]\n","\n","        # self.fusion_of_context = [3]\n","        # self.visual_transformer = TransformerEncoder(d_model = VISUAL_DIM,\n","        #                                              n_layers = 4,\n","        #                                              n_heads=8,\n","        #                                              d_ff=VISUAL_DIM\n","        #                                              )\n","        # self.acoustic_transformer = TransformerEncoder(d_model = ACOUSTIC_DIM,\n","        #                                                n_layers=4,\n","        #                                                n_heads=2,\n","        #                                                d_ff=ACOUSTIC_DIM)\n","\n","        # self.MAF_layer3 = MAF(dim_model=embed_dim,\n","        #                      dropout_rate=0.2)\n","        \n","        self.MAF_layer4 = MAF_acoustic(dim_model=embed_dim,\n","                             dropout_rate=0.2)\n","        \n","        self.MAF_layer5 = MAF_visual(dim_model=embed_dim,\n","                             dropout_rate=0.2)\n","        \n","        # self.context_encoder = ContextEncoder(config)\n","\n","        # self.classification = nn.Linear(embed_dim, 2)\n","\n","       \n","\n","    def forward(self,\n","            input_ids = None,\n","            attention_mask = None,\n","            # context_input_ids = None,\n","            # context_attention_mask = None,\n","            acoustic_input = None,\n","            visual_input = None,\n","            head_mask = None,\n","            inputs_embeds = None,\n","            output_attentions = None,\n","            output_hidden_states  = None,\n","            return_dict = None):\n","\n","            # print(\"Input ids shape : \", input_ids.shape)\n","            output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions\n","\n","            output_hidden_states = (\n","                output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states\n","            )\n","\n","            return_dict = return_dict if return_dict is not None else self.config.use_return_dict\n","\n","            if input_ids is not None and inputs_embeds is not None:\n","                raise ValueError(\"You can't specify both input_ids and inputs_embeds at the same time\")\n","            elif input_ids is not None:\n","                input_shape = input_ids.size()\n","                input_ids = input_ids.view(-1, input_shape[-1])\n","\n","            elif inputs_embeds is not None:\n","                input_shape = inputs_embeds.size()[:-1]\n","            else:\n","                raise ValueError(\"You have to specify either input_ids or input_embeds\")\n","\n","\n","            if inputs_embeds is None:\n","                # print(\"Input ids shape : \", input_ids.shape)\n","                inputs_embeds = self.embed_tokens(input_ids) * self.embed_scale\n","\n","            # print(\"Input shape type : \", type(input_shape))\n","            # print(\"Input shape : \", input_shape)\n","            input_shape = torch.tensor(input_shape)\n","            # print(\"Input shape type : \", type(input_shape))\n","            # print(\"Input shape : \", input_shape)\n","            # print(\"Input shape : \", input_shape.shape)\n","            embed_pos = self.embed_positions(input_ids)\n","            # embed_pos = self.embed_positions(input_shape)\n","\n","\n","            hidden_states = inputs_embeds + embed_pos   \n","            hidden_states = self.layernorm_embedding(hidden_states)\n","            hidden_states = F.dropout(hidden_states, p = self.dropout, training=self.training)\n","\n","            # print(\"attention mask shape 3 : \", attention_mask.shape)\n","            if attention_mask is not None:\n","                attention_mask =  _expand_mask(attention_mask, inputs_embeds.dtype)\n","\n","            # print(\"attention mask shape 4 : \", attention_mask.shape)\n","            encoder_states = () if output_hidden_states else None\n","            all_attentions = () if output_attentions else None\n","\n","            if head_mask is not None:\n","                assert head_mask.size()[0] == (\n","                    len(self.layers)\n","                ), f\"The head mask should be specified for {len(self.layers)} layers, but it is for {head_mask.size()[0]}.\"\n","\n","            for idx, encoder_layer in enumerate(self.layers):\n","                # print(\"============Idx : \", idx)\n","\n","                # if idx in self.fusion_at_layer3:\n","                #     # print(\"Acoustic input shape (B) : \", acoustic_input)\n","                #     # acoustic_input = self.acoustic_transformer(acoustic_input)[-1]   \n","                #     # print(\"Acoustic input shape (C) : \", acoustic_input)\n","\n","                #     # visual_input = self.visual_transformer(visual_input)[-1]\n","                #     # print(\"====Idx inside fusion at layer :\", idx)\n","                #     hidden_states = self.MAF_layer3(text_input = hidden_states,\n","                #                                    acoustic_context = acoustic_input,\n","                #                                    visual_context = visual_input) \n","\n","                # if idx in self.fusion_of_context:\n","                  \n","                #   hidden_states = self.context_encoder(hidden_states = hidden_states, context_input_ids = context_input_ids, context_attention_mask = context_attention_mask)\n","                  \n","\n","                if idx in self.fusion_at_layer4:\n","                    # print(\"Acoustic input shape (B) : \", acoustic_input)\n","                    # acoustic_input = self.acoustic_transformer(acoustic_input)[-1]   \n","                    # print(\"Acoustic input shape (C) : \", acoustic_input)\n","\n","                    # visual_input = self.visual_transformer(visual_input)[-1]\n","                    # print(\"====Idx inside fusion at layer :\", idx)\n","\n","                    \n","                    \n","                    hidden_states = self.MAF_layer4(text_input = hidden_states,\n","                                                   acoustic_context = acoustic_input\n","                                                   )\n","                if idx in self.fusion_at_layer5:\n","                    # print(\"Acoustic input shape (B) : \", acoustic_input)\n","                    # acoustic_input = self.acoustic_transformer(acoustic_input)[-1]   \n","                    # print(\"Acoustic input shape (C) : \", acoustic_input)\n","\n","                    # visual_input = self.visual_transformer(visual_input)[-1]\n","                    # print(\"====Idx inside fusion at layer :\", idx)\n","                    hidden_states = self.MAF_layer5(text_input = hidden_states,\n","                                                   visual_context = visual_input)     \n","\n","                if output_hidden_states:\n","                    encoder_states = encoder_states + (hidden_states,)\n","\n","                dropout_probability = random.uniform(0,1)   \n","\n","                if self.training and (dropout_probability < self.layerdrop):\n","                    layer_outputs = (None, None)\n","\n","                else:\n","                    if self.gradient_checkpointing and self.training:\n","\n","                        def create_custom_forward(module):\n","                            def custom_forward(*inputs):\n","                                return module(*inputs, output_attentions)\n","\n","                            return custom_forward\n","\n","                        layer_outputs = torch.utils.checkpoint.checkpoint(\n","                            create_custom_forward(encoder_layer),\n","                            hidden_states,\n","                            attention_mask,\n","                            (head_mask[idx] if head_mask is not None else None),\n","                        )        \n","\n","                    else:\n","                        # print(\"Checking Attention mask shape : \", attention_mask.shape)\n","                        layer_outputs = encoder_layer(\n","                            hidden_states,\n","                            attention_mask,\n","                            layer_head_mask = (head_mask[idx] if head_mask is not None else None),\n","                            output_attentions = output_attentions\n","                        )    \n","\n","                    hidden_states = layer_outputs[0]\n","\n","                if output_attentions:\n","                    all_attentions  = all_attentions + (layer_outputs[1],)\n","\n","            if output_hidden_states:\n","                encoder_states = encoder_states + (hidden_states,)\n","\n","            if not return_dict:\n","                return tuple(v for v in [hidden_states, encoder_states, all_attentions] if v is not None)\n","\n","            return BaseModelOutput(\n","                last_hidden_state=hidden_states, hidden_states=encoder_states, attentions=all_attentions\n","            )    \n","\n","            # print(\"Hidden states shape : \", hidden_states)\n","\n","            # cls = hidden_states.permute(1,0,2)"]},{"cell_type":"code","execution_count":29,"metadata":{"id":"6F_nfly0HgKR","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MultimodalBartModel(BartPretrainedModel):\n","    def __init__(self, config: BartConfig):\n","        super().__init__(config)\n","\n","        padding_idx, vocab_size = config.pad_token_id, config.vocab_size\n","        self.shared = nn.Embedding(vocab_size, config.d_model, padding_idx)\n","\n","        self.encoder = MultiModalBartEncoder(config, self.shared)\n","        self.decoder = BartDecoder(config, self.shared)\n","\n","        self.init_weights()\n","\n","    def get_input_embeddings(self):\n","        return self.shared\n","\n","    def set_input_embeddings(self, value):\n","        self.shared = value\n","        self.encoder.embed_tokens = self.shared\n","        self.decoder.embed_tokens = self.shared\n","\n","    def get_encoder(self):\n","        return self.encoder\n","\n","\n","    def get_decoder(self):\n","        return self.decoder\n","\n","    def forward(\n","        self, \n","        input_ids = None,\n","        attention_mask = None,\n","        # context_input_ids = None,\n","        # context_attention_mask = None,\n","        acoustic_input = None,\n","        visual_input = None,\n","        decoder_input_ids = None,\n","        decoder_attention_mask = None,\n","        head_mask = None,\n","        decoder_head_mask = None,\n","        cross_attn_head_mask = None,\n","        encoder_outputs = None,\n","        past_key_values = None,\n","        inputs_embeds = None,\n","        decoder_inputs_embeds = None,\n","        use_cache = None,\n","        output_attentions = None,\n","        output_hidden_states = None,\n","        return_dict = None\n","    ):\n","\n","        if decoder_input_ids is None and decoder_inputs_embeds is None:\n","            decoder_input_ids = shift_tokens_right(\n","                input_ids, self.config.pad_token_id, self.config.decoder_start_token_id\n","\n","            )                                \n","\n","        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions\n","        output_hidden_states = (\n","            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states\n","        )    \n","        use_cache = use_cache if use_cache is not None else self.config.use_cache\n","        return_dict = return_dict if return_dict is not None else self.config.use_return_dict\n","\n","        # print(\"attention mask shape 2 : \", attention_mask.shape)\n","\n","        if encoder_outputs is None:\n","            encoder_outputs = self.encoder(\n","                input_ids = input_ids,\n","                attention_mask = attention_mask,\n","                # context_input_ids = context_input_ids,\n","                # context_attention_mask = context_attention_mask,\n","                acoustic_input = acoustic_input,\n","                visual_input = visual_input,\n","                head_mask = head_mask,\n","                inputs_embeds = inputs_embeds,\n","                output_attentions = output_attentions,\n","                output_hidden_states = output_hidden_states,\n","                return_dict = return_dict\n","            )\n","\n","        elif return_dict and not isinstance(encoder_outputs, BaseModelOutput):\n","            encoder_outputs = BaseModelOutput(\n","                last_hidden_state=encoder_outputs[0],\n","                hidden_states=encoder_outputs[1] if len(encoder_outputs) > 1 else None,\n","                attentions = encoder_outputs[2] if len(encoder_outputs) > 2 else None\n","            )    \n","\n","        decoder_outputs = self.decoder(\n","            input_ids = decoder_input_ids,\n","            attention_mask = decoder_attention_mask,\n","            encoder_hidden_states = encoder_outputs[0],\n","            encoder_attention_mask = attention_mask,\n","            head_mask = decoder_head_mask,\n","            cross_attn_head_mask = cross_attn_head_mask,\n","            past_key_values = past_key_values,\n","            inputs_embeds = decoder_inputs_embeds,\n","            use_cache = use_cache,\n","            output_attentions = output_attentions,\n","            output_hidden_states = output_hidden_states,\n","            return_dict = return_dict\n","        )    \n","\n","        if not return_dict:\n","            return decoder_outputs + encoder_outputs\n","\n","        return Seq2SeqModelOutput(\n","            last_hidden_state=decoder_outputs.last_hidden_state,\n","            past_key_values=decoder_outputs.past_key_values,\n","            decoder_hidden_states=decoder_outputs.hidden_states,\n","            decoder_attentions=decoder_outputs.attentions,\n","            cross_attentions=decoder_outputs.cross_attentions,\n","            encoder_last_hidden_state=encoder_outputs.last_hidden_state,\n","            encoder_hidden_states=encoder_outputs.hidden_states,\n","            encoder_attentions=encoder_outputs.attentions\n","\n","        )"]},{"cell_type":"code","execution_count":30,"metadata":{"id":"pWEHZicLrtxY","executionInfo":{"status":"ok","timestamp":1684514971230,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class MultiHeadSelfAttention(torch.nn.Module):\n","#   def __init__(self, config):\n","#     super(MultiHeadSelfAttention, self).__init__()\n","\n","#     self.num_attention_heads = 12\n","#     self.attention_head_size = config.d_model//12\n","#     self.embed_dim = config.d_model\n","#     self.dropout = config.dropout\n","\n","#     self.query = nn.Linear(config.d_model, config.d_model)\n","#     self.key = nn.Linear(config.d_model, config.d_model)\n","#     self.value = nn.Linear(config.d_model, config.d_model)\n","#     self.out_proj = nn.Linear(config.d_model, config.d_model)\n","\n","#   def forward(self, hidden_states, attention_mask):\n","#     query_states = self.query(hidden_states)\n","#     key_states = self.key(hidden_states)\n","#     value_states = self.value(hidden_states)\n","\n","#     batch_size = hidden_states.shape[0]\n","#     seq_len = hidden_states.shape[1]\n","\n","#     query_states =  query_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","#     key_states =  key_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","#     value_states = value_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","\n","#     query_states = query_states.view(batch_size*self.num_attention_heads, -1, self.attention_head_size)\n","#     key_states  = key_states.view(batch_size*self.num_attention_heads, -1, self.attention_head_size)\n","#     value_states = value_states.view(batch_size*self.num_attention_heads, -1, self.attention_head_size)\n","\n","#     attn_weights = torch.bmm(query_states, key_states.transpose(1,2))\n","\n","#     attn_weights = attn_weights.view(batch_size, self.num_attention_heads, attn_weights.shape[1], attn_weights.shape[2])\n","#     attn_weights = attn_weights/torch.sqrt(torch.tensor(self.attention_head_size))\n","\n","#     attn_weights_masked = attn_weights.masked_fill(\n","#         attention_mask[:, None, None, :] == False, -1e16\n","#     )\n","\n","#     attn_weights = torch.nn.functional.softmax(attn_weights_masked, dim = -1)\n","\n","#     attn_weights = torch.nn.functional.dropout(attn_weights, p = self.dropout, training = self.training)\n","\n","#     attn_weights = attn_weights.view(batch_size*self.num_attention_heads, seq_len, seq_len)\n","\n","\n","    \n","#     attn_output = torch.bmm(attn_weights, value_states)\n","\n","#     attn_output = attn_output.view(batch_size, self.num_attention_heads, seq_len, self.attention_head_size)\n","#     attn_output = attn_output.transpose(1,2)\n","\n","#     attn_output = attn_output.reshape(batch_size, seq_len, self.embed_dim)\n","\n","#     attn_out = self.out_proj(attn_output)"]},{"cell_type":"code","execution_count":31,"metadata":{"id":"TNYKv7lfrFq0","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class SelfAttention(torch.nn.Module):\n","#   def __init__(self):\n","#     super(SelfAttention, self).__init__()\n","#     self.num_attention_heads = 1\n","#     self.attention_head_size = 768\n","#     self.embed_dim = 768\n","\n","#     self.query = torch.nn.Linear(768, 768, bias = False)\n","#     self.key = torch.nn.Linear(768, 768, bias = False)\n","#     self.value = torch.nn.Linear(768, 768, bias = False)\n","\n","#     self.out_proj = torch.nn.Linear(768, 768, bias = False)\n","\n","#   def forward(self, hidden_states, attention_mask)  :\n","#     query_states = self.query(hidden_states)\n","#     key_states = self.key(hidden_states)\n","#     value_states = self.value(hidden_states)\n","\n","#     batch_size = hidden_states.shape[0]\n","#     seq_len = hidden_states.shape[1]\n","\n","#     query_states = query_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","#     key_states = key_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","#     value_states = value_states.view(batch_size, seq_len, self.num_attention_heads, self.attention_head_size).transpose(1,2).contiguous()\n","\n","#     attn_weights = torch.matmul(query_states, key_states.transpose(2,3))\n","#     attn_weights = attn_weights/torch.sqrt(torch.tensor(self.attention_head_size))\n","\n","#     attn_weights_masked = attn_weights.masked_fill(\n","#         attention_mask[:, None, None, :] == False, -1e16\n","#     )\n","\n","#     attn_weights = torch.nn.functional.softmax(attn_weights_masked, dim = -1)\n","\n","#     attn_output = torch.matmul(attn_weights, value_states)\n","\n","#     attn_output = attn_output.transpose(1,2).contiguous()\n","\n","#     attn_output = attn_output.view(batch_size, seq_len, -1)\n","\n","#     attn_out = self.out_proj(attn_output)\n","\n","#     return attn_out\n","\n","  "]},{"cell_type":"code","execution_count":32,"metadata":{"id":"VO1KdKseHlkn","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MultimodalBartClassification(nn.Module):\n","    def __init__(\n","        self, \n","        input_dim: int,\n","        inned_dim: int,\n","        num_classes: int,\n","        pooler_dropout: float\n","    ):\n","        super().__init__()\n","        self.dense = nn.Linear(input_dim, inned_dim)\n","        self.dropout = nn.Dropout(p = pooler_dropout)\n","        self.out_proj = nn.Linear(inned_dim, num_classes)\n","\n","    def forward(self, hidden_states):\n","        hidden_states = self.dropout(hidden_states)\n","        hidden_states = self.dense(hidden_states)\n","        hidden_states = torch.tanh(hidden_states)\n","        temp_hid = hidden_states\n","        hidden_states = self.dropout(hidden_states)\n","        hidden_states = self.out_proj(hidden_states)\n","        return hidden_states, temp_hid      \n","\n","\n","\n","\n","class MultimodalBartForSequenceClassification(BartPretrainedModel):\n","    _keys_to_ignore_on_load_missing = [\"encoder.embed_tokens.weight\", \"decoder.embed_tokens.weight\"]\n","\n","    def __init__(self, config: BartConfig, **kwargs):\n","        super().__init__(config, **kwargs)\n","        self.model = MultimodalBartModel(config)\n","        # self.bigru = torch.nn.GRU(input_size = 768, hidden_size = 384, num_layers = 1, batch_first = True, dropout = 0.1,  bidirectional = True)\n","        # self.custom_attention = SelfAttention()\n","        self.classification_head = MultimodalBartClassification(\n","            config.d_model,\n","            config.d_model,\n","            2,\n","            config.classifier_dropout\n","        )\n","        self.model._init_weights(self.classification_head.dense)\n","        self.model._init_weights(self.classification_head.out_proj)\n","\n","    def forward(\n","        self,\n","        input_ids: torch.LongTensor = None,\n","        attention_mask: Optional[torch.tensor] = None,\n","        # context_input_ids : torch.LongTensor = None,\n","        # context_attention_mask : Optional[torch.tensor] = None,\n","        acoustic_input = None,\n","        visual_input = None,\n","        decoder_input_ids: Optional[torch.LongTensor] = None,\n","        decoder_attention_mask: Optional[torch.LongTensor] = None,\n","        head_mask: Optional[torch.Tensor] = None,\n","        decoder_head_mask: Optional[torch.Tensor] = None,\n","        cross_attn_head_mask: Optional[torch.Tensor] = None,\n","        encoder_outputs: Optional[List[torch.FloatTensor]] = None,\n","        inputs_embeds: Optional[torch.FloatTensor] = None,\n","        decoder_inputs_embeds: Optional[torch.FloatTensor] = None,\n","        labels: Optional[torch.LongTensor] = None,\n","        use_cache: Optional[bool] = None,\n","        output_attentions: Optional[bool] = None,\n","        output_hidden_states: Optional[bool] = None,\n","        return_dict: Optional[bool] = None\n","    ) -> Union[Tuple, Seq2SeqSequenceClassifierOutput]:\n","        r\"\"\"\n","        labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):\n","            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,\n","            config.num_labels - 1]`. If `config.num_labels > 1` a classification loss is computed (Cross-Entropy).\n","        \"\"\"\n","\n","        return_dict = return_dict if return_dict is not None else self.config.use_return_dict\n","        if labels is not None:\n","            use_cache = False\n","\n","        if input_ids is None and inputs_embeds is not None:\n","            raise NotImplementedError(\n","                f\"Passing input embeddings is currently not supported for {self.__class__.__name__}\"\n","            )\n","\n","        # print(\"attention mask shape 1 : \", attention_mask.shape )\n","        outputs = self.model(\n","            input_ids,\n","            attention_mask = attention_mask,\n","            # context_input_ids = context_input_ids,\n","            # context_attention_mask = context_attention_mask,\n","            acoustic_input = acoustic_input,\n","            visual_input = visual_input,\n","            decoder_input_ids = decoder_input_ids,\n","            decoder_attention_mask = decoder_attention_mask,\n","            head_mask = head_mask,\n","            decoder_head_mask = decoder_head_mask,\n","            cross_attn_head_mask = cross_attn_head_mask,\n","            encoder_outputs = encoder_outputs,\n","            inputs_embeds = inputs_embeds,\n","            decoder_inputs_embeds = decoder_inputs_embeds,\n","            use_cache = use_cache,\n","            output_attentions = output_attentions,\n","            output_hidden_states = output_hidden_states,\n","            return_dict = return_dict\n","\n","        )    \n","\n","        hidden_states = outputs[0]\n","\n","        eos_mask = input_ids.eq(self.config.eos_token_id).to(hidden_states.device)\n","\n","        if len(torch.unique_consecutive(eos_mask.sum(1))) > 1:\n","            raise ValueError(\"All examples must have same number of <eos> tokens\")\n","\n","        sentence_representation = hidden_states[eos_mask, :].view(hidden_states.size(0), -1, hidden_states.size(-1))[\n","            :, -1, :\n","        ]    \n","\n","        logits, temp_hid = self.classification_head(sentence_representation)\n","\n","        # print(\"hidden_states shape : \", hidden_states.shape)\n","\n","        # gru_out, _ = self.bigru(hidden_states)\n","\n","        # print(\"gru out shape : \", gru_out.shape)\n","\n","      \n","\n","        # attn_output = self.custom_attention(gru_out, attention_mask)\n","\n","        # print(\"attention_output shape : \", attn_output.shape)\n","\n","        # logits = self.classification_head(attn_output[:, -1, :])\n","        # logits = self.classification_head(gru_out[:, -1, :])\n","\n","        \n","\n","        loss = None\n","\n","        loss_fct = CrossEntropyLoss()\n","        # print(\"logits shape : \", logits.shape)\n","        loss = loss_fct(logits.view(-1, 2), labels.view(-1))\n","        # print(\"Loss : \", loss)\n","        if not return_dict:\n","            output = (logits,) + outputs[1:]\n","            return ((loss,) + output) if loss is not None else output\n","\n","\n","        return Seq2SeqSequenceClassifierOutput(\n","            loss=loss,\n","            logits=logits,\n","            past_key_values=outputs.past_key_values,\n","            decoder_hidden_states=outputs.decoder_hidden_states,\n","            decoder_attentions = outputs.decoder_attentions,\n","            cross_attentions=outputs.cross_attentions,\n","            encoder_last_hidden_state=outputs.encoder_last_hidden_state,\n","            encoder_hidden_states=outputs.encoder_hidden_states,\n","            encoder_attentions=outputs.encoder_attentions\n","        ), sentence_representation , hidden_states, temp_hid"]},{"cell_type":"markdown","metadata":{"id":"klRdsBKNjU_l"},"source":["#Shared Sarcasm"]},{"cell_type":"code","execution_count":33,"metadata":{"id":"aVU7tDvpgWAx","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class MultimodalSharedSarcasmClassification(nn.Module):\n","#     def __init__(\n","#         self, \n","#         input_dim: int,\n","#         inned_dim: int,\n","#         # inned_dim2: int,\n","#         num_classes: int,\n","#         pooler_dropout: float\n","#     ):\n","#         super().__init__()\n","#         self.dense = nn.Linear(input_dim, inned_dim)\n","#         self.dropout = nn.Dropout(p = pooler_dropout)\n","#         # self.dense2 = nn.Linear(inned_dim, inned_dim2)\n","#         self.out_proj = nn.Linear(inned_dim, num_classes)\n","\n","#     def forward(self, hidden_states):\n","#         hidden_states = self.dropout(hidden_states)\n","#         hidden_states = self.dense(hidden_states)\n","#         hidden_states = torch.tanh(hidden_states)\n","#         hidden_states = self.dropout(hidden_states)\n","#         # hidden_states = self.dense2(hidden_states)\n","#         # hidden_states = torch.tanh(hidden_states)\n","#         hidden_states = self.out_proj(hidden_states)\n","#         return hidden_states    "]},{"cell_type":"code","execution_count":34,"metadata":{"id":"kLaKAidQgRQD","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class SarcasmShared(torch.nn.Module):\n","#   def __init__(self):\n","#     super(SarcasmShared, self).__init__()\n","#     # self.gate = torch.nn.Linear(768*2, 768)\n","#     # self.attention_layer = nn.MultiheadAttention(embed_dim=768,\n","#     #                                                  num_heads = 1,\n","#     #                                                 #  dropout = self.dropout_rate,\n","#     #                                                  bias = True,\n","#     #                                                 add_zero_attn=False,\n","#     #                                                 batch_first=True,\n","#     #                                                 device=DEVICE   \n","#     #     )\n","#     self.alphas_a = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","#     self.alphas_v = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","#     self.alphas_c = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","\n","#     self.transform = torch.nn.Linear(768,768)\n","#     self.classifier = torch.nn.Linear(768, 12)\n","\n","#   def forward(self, sarcasm_logits1, intent_logits1, sarcasm_logits2, intent_logits2):\n","#     batch = sarcasm_logits1.shape[0]\n","#     central_rep =torch.zeros((batch,768)).to(device)\n","#     xsum1 = sarcasm_logits1*self.alphas_a[0].expand_as(sarcasm_logits1) + central_rep*self.alphas_c[0].expand_as(central_rep) + intent_logits1*self.alphas_v[0].expand_as(intent_logits1)\n","#     central_rep = self.transform(xsum1)\n","#     xsum2 = sarcasm_logits2*self.alphas_a[1].expand_as(sarcasm_logits2) + central_rep*self.alphas_c[1].expand_as(central_rep) + intent_logits2*self.alphas_v[1].expand_as(intent_logits2)\n","#     central_out = self.classifier(xsum2)\n","\n","    \n","    \n","#     # shared_out = torch.cat([sarcasm_logits, intent_logits], dim = -1)\n","\n","#     # shared_out = self.classification_head(shared_out)\n","\n","#     # return shared_out\n","\n","#     return central_out\n"]},{"cell_type":"markdown","metadata":{"id":"ELrFVNETjSlW"},"source":["#Shared Intent"]},{"cell_type":"code","execution_count":35,"metadata":{"id":"ek48By4ngYhU","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class MultimodalSharedIntentClassification(nn.Module):\n","#     def __init__(\n","#         self, \n","#         input_dim: int,\n","#         inned_dim: int,\n","#         num_classes: int,\n","#         pooler_dropout: float\n","#     ):\n","#         super().__init__()\n","#         self.dense = nn.Linear(input_dim, inned_dim)\n","#         self.dropout = nn.Dropout(p = pooler_dropout)\n","#         self.out_proj = nn.Linear(inned_dim, num_classes)\n","\n","#     def forward(self, hidden_states):\n","#         hidden_states = self.dropout(hidden_states)\n","#         hidden_states = self.dense(hidden_states)\n","#         hidden_states = torch.tanh(hidden_states)\n","#         hidden_states = self.dropout(hidden_states)\n","#         hidden_states = self.out_proj(hidden_states)\n","#         return hidden_states    "]},{"cell_type":"code","execution_count":36,"metadata":{"id":"xvi__Q1wgZzU","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class IntentShared(torch.nn.Module):\n","#   def __init__(self):\n","#     super(IntentShared, self).__init__()\n","#     self.classification_head = MultimodalSharedIntentClassification(\n","#         768*2,\n","#         768,\n","#         13,\n","#         0.1\n","#     )\n","\n","#   def forward(self, sarcasm_logits, intent_logits):\n","#     shared_out = torch.cat([sarcasm_logits, intent_logits], dim = -1)\n","\n","#     shared_out = self.classification_head(shared_out)\n","\n","#     return shared_out\n","\n"]},{"cell_type":"code","source":["# class IntentShared(torch.nn.Module):\n","#   def __init__(self):\n","#     super(IntentShared, self).__init__()\n","#     # self.classification_head = MultimodalSharedIntentClassification(\n","#     #     768,\n","#     #     768,\n","#     #     13,\n","#     #     0.1\n","#     # )\n","#     self.alphas_a = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","#     self.alphas_v = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","#     self.alphas_c = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","\n","#     self.transform = torch.nn.Linear(768,768)\n","#     self.classifier = torch.nn.Linear(768, 12)\n","\n","#   def forward(self, sarcasm_logits1, intent_logits1, sarcasm_logits2, intent_logits2):\n","#     batch = sarcasm_logits1.shape[0]\n","#     central_rep =torch.zeros((batch,768)).to(device)\n","#     xsum1 = sarcasm_logits1*self.alphas_a[0].expand_as(sarcasm_logits1) + central_rep*self.alphas_c[0].expand_as(central_rep) + intent_logits1*self.alphas_v[0].expand_as(intent_logits1)\n","#     central_rep = self.transform(xsum1)\n","#     xsum2 = sarcasm_logits2*self.alphas_a[1].expand_as(sarcasm_logits2) + central_rep*self.alphas_c[1].expand_as(central_rep) + intent_logits2*self.alphas_v[1].expand_as(intent_logits2)\n","#     central_out = self.classifier(xsum2)\n","\n","    \n","    \n","#     # shared_out = torch.cat([sarcasm_logits, intent_logits], dim = -1)\n","\n","#     # shared_out = self.classification_head(shared_out)\n","\n","#     # return shared_out\n","\n","#     return central_out\n","\n"],"metadata":{"id":"59onId0_Tqn4","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":37,"outputs":[]},{"cell_type":"code","execution_count":38,"metadata":{"id":"xMrNm9srjN4h","executionInfo":{"status":"ok","timestamp":1684514971231,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class IntentShared(torch.nn.Module):\n","  def __init__(self):\n","    super(IntentShared, self).__init__()\n","    # self.gate = torch.nn.Linear(768*2, 768)\n","    self.gate = torch.nn.Linear(768*2, 768)\n","    # self.attention_layer = nn.MultiheadAttention(embed_dim=768,\n","    #                                                  num_heads = 1,\n","    #                                                 #  dropout = self.dropout_rate,\n","    #                                                  bias = True,\n","    #                                                 add_zero_attn=False,\n","    #                                                 batch_first=True,\n","    #                                                 device=DEVICE   \n","    #     )\n","    # self.alphas_a = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","    # self.alphas_v = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","    # self.alphas_c = nn.ParameterList([torch.nn.Parameter(torch.rand(1)) for i in range(2)])\n","\n","    # self.transform = torch.nn.Linear(768,768)\n","    self.classifier = torch.nn.Linear(768, 12)\n","\n","  def forward(self, sarcasm_logits1, intent_logits1, sarcasm_logits2, intent_logits2):\n","    # batch = sarcasm_logits1.shape[0]\n","    # central_rep =torch.zeros((batch,768)).to(device)\n","    # xsum1 = sarcasm_logits1*self.alphas_a[0].expand_as(sarcasm_logits1) + central_rep*self.alphas_c[0].expand_as(central_rep) + intent_logits1*self.alphas_v[0].expand_as(intent_logits1)\n","    # central_rep = self.transform(xsum1)\n","    # xsum2 = sarcasm_logits2*self.alphas_a[1].expand_as(sarcasm_logits2) + central_rep*self.alphas_c[1].expand_as(central_rep) + intent_logits2*self.alphas_v[1].expand_as(intent_logits2)\n","    # central_out = self.classifier(xsum2)\n","\n","    \n","    \n","    shared_out = torch.cat([sarcasm_logits2, intent_logits2], dim = -1)\n","\n","    shared_out = self.gate(shared_out)\n","\n","    shared_out = torch.relu(shared_out)\n","\n","    shared_out = self.classifier(shared_out)\n","\n","    return shared_out\n","\n","    # return central_out\n","\n"]},{"cell_type":"markdown","metadata":{"id":"WeHRb6R1jO8c"},"source":["#Intent Only"]},{"cell_type":"code","execution_count":39,"metadata":{"id":"C-iaNNLQgiwO","executionInfo":{"status":"ok","timestamp":1684514971232,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class IntentContextAwareAttention(nn.Module):\n","\n","    def __init__(self,\n","                 dim_model : int,\n","                 dim_context : int,\n","                 dropout_rate : Optional[float] = 0.0 ):\n","\n","        super(IntentContextAwareAttention, self).__init__()\n","\n","        self.dim_model = dim_model\n","        self.dim_context = dim_context\n","        self.dropout_rate = dropout_rate\n","        self.attention_layer = nn.MultiheadAttention(embed_dim=self.dim_model,\n","                                                     num_heads = 1,\n","                                                     dropout = self.dropout_rate,\n","                                                     bias = True,\n","                                                    add_zero_attn=False,\n","                                                    batch_first=True,\n","                                                    device=DEVICE   \n","        )\n","\n","        self.u_k = nn.Linear(self.dim_context, self.dim_model, bias = False)\n","        self.w1_k = nn.Linear(self.dim_model, 1, bias=False)\n","        self.w2_k = nn.Linear(self.dim_model, 1, bias=False)\n","\n","        self.u_v = nn.Linear(self.dim_context, self.dim_model, bias=False)\n","        self.w1_v = nn.Linear(self.dim_model, 1, bias = False)\n","        self.w2_v = nn.Linear(self.dim_model, 1, bias = False)\n","\n","    def forward(self, q, k, v, context):\n","\n","        # print(\"Context shape : \", context.shape)\n","        # print(\"Dim context : \", self.dim_context, \" : Dim model : \", self.dim_model)\n","        key_context = self.u_k(context)\n","        # print(\"Context shape below key context : \", key_context.shape)\n","        value_context = self.u_v(context)\n","\n","        lambda_k = F.sigmoid(self.w1_k(k) + self.w2_k(key_context))\n","        lambda_v = F.sigmoid(self.w1_v(v) + self.w2_v(value_context))\n","\n","        k_cap = (1-lambda_k) * k + (lambda_k) * key_context\n","        v_cap = (1-lambda_v) * v + (lambda_v) * value_context\n","\n","        attention_output, _ = self.attention_layer(query = q,\n","                                                   key = k_cap,\n","                                                   value = v_cap)\n","  \n","        return attention_output                                     \n","\n"]},{"cell_type":"code","execution_count":40,"metadata":{"id":"9t9Vb8cdidti","executionInfo":{"status":"ok","timestamp":1684514971232,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MAF_Intent_acoustic(nn.Module):\n","    def __init__(self, \n","                dim_model,\n","                dropout_rate):\n","        super(MAF_Intent_acoustic, self).__init__()\n","        self.dropout_rate = dropout_rate\n","\n","        self.acoustic_context_transform = nn.Linear(ACOUSTIC_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","        # self.visual_context_transform = nn.Linear(VISUAL_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","\n","        self.acoustic_context_attention = IntentContextAwareAttention(dim_model=dim_model,\n","                                                                dim_context=ACOUSTIC_DIM,\n","                                                                dropout_rate=dropout_rate)\n","\n","        # self.visual_context_attention = ContextAwareAttention(dim_model=dim_model,\n","        #                                                     dim_context=VISUAL_DIM,\n","        #                                                     dropout_rate=dropout_rate)\n","\n","        self.acoustic_gate = nn.Linear(2*dim_model, dim_model)\n","        # self.visual_gate = nn.Linear(2*dim_model, dim_model)\n","        self.dropout_layer = nn.Dropout(dropout_rate)\n","        self.final_layer_norm = nn.LayerNorm(dim_model)\n","\n","    def forward(self,\n","                text_input,\n","                acoustic_context):\n","\n","        # print(\"Acoustic context shape (A) : \", acoustic_context.shape)        \n","\n","        acoustic_context = acoustic_context.permute(0,2,1)\n","        acoustic_context = self.acoustic_context_transform(acoustic_context.float())\n","        acoustic_context = acoustic_context.permute(0,2,1)\n","\n","        audio_out = self.acoustic_context_attention(q=text_input,\n","                                                    k=text_input,\n","                                                    v=text_input,\n","                                                    context=acoustic_context)\n","        # print(\"Audio out (A) : \", audio_out.shape) \n","\n","        # print(\"Visual context shape : \", visual_context.shape)\n","        # visual_context = visual_context.permute(0,2,1)\n","        # visual_context = self.visual_context_transform(visual_context.float())\n","        # visual_context = visual_context.permute(0,2,1)\n","        \n","        # video_out = self.visual_context_attention(q=text_input,\n","        #                                             k=text_input,\n","        #                                             v=text_input,\n","        #                                             context=visual_context)\n","\n","        # print(\"Video out shape : \", video_out.shape)\n","        # print(\"Text input shape : \", text_input.shape)\n","        weight_a = F.sigmoid(self.acoustic_gate(torch.cat([text_input, audio_out], dim=-1)))\n","        # weight_v = F.sigmoid(self.visual_gate(torch.cat([text_input, video_out], dim=-1)))\n","\n","        # output = self.final_layer_norm(text_input + weight_a * audio_out + weight_v * video_out)\n","\n","        output = self.final_layer_norm(text_input + weight_a * audio_out)\n","\n","        return output"]},{"cell_type":"code","execution_count":41,"metadata":{"id":"f3-SvUJoiffh","executionInfo":{"status":"ok","timestamp":1684514971232,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MAF_Intent_visual(nn.Module):\n","    def __init__(self, \n","                dim_model,\n","                dropout_rate):\n","        super(MAF_Intent_visual, self).__init__()\n","        self.dropout_rate = dropout_rate\n","\n","        # self.acoustic_context_transform = nn.Linear(ACOUSTIC_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","        self.visual_context_transform = nn.Linear(VISUAL_MAX_LEN, SOURCE_MAX_LEN, bias = False)\n","\n","        # self.acoustic_context_attention = ContextAwareAttention(dim_model=dim_model,\n","        #                                                         dim_context=ACOUSTIC_DIM,\n","        #                                                         dropout_rate=dropout_rate)\n","\n","        self.visual_context_attention = IntentContextAwareAttention(dim_model=dim_model,\n","                                                            dim_context=VISUAL_DIM,\n","                                                            dropout_rate=dropout_rate)\n","\n","        # self.acoustic_gate = nn.Linear(2*dim_model, dim_model)\n","        self.visual_gate = nn.Linear(2*dim_model, dim_model)\n","        self.dropout_layer = nn.Dropout(dropout_rate)\n","        self.final_layer_norm = nn.LayerNorm(dim_model)\n","\n","    def forward(self,\n","                text_input,\n","                visual_context):\n","\n","        # print(\"Acoustic context shape (A) : \", acoustic_context.shape)        \n","\n","        # acoustic_context = acoustic_context.permute(0,2,1)\n","        # acoustic_context = self.acoustic_context_transform(acoustic_context.float())\n","        # acoustic_context = acoustic_context.permute(0,2,1)\n","\n","        # audio_out = self.acoustic_context_attention(q=text_input,\n","        #                                             k=text_input,\n","        #                                             v=text_input,\n","        #                                             context=acoustic_context)\n","        # print(\"Audio out (A) : \", audio_out.shape) \n","\n","        # print(\"Visual context shape : \", visual_context.shape)\n","        visual_context = visual_context.permute(0,2,1)\n","        visual_context = self.visual_context_transform(visual_context.float())\n","        visual_context = visual_context.permute(0,2,1)\n","        \n","        video_out = self.visual_context_attention(q=text_input,\n","                                                    k=text_input,\n","                                                    v=text_input,\n","                                                    context=visual_context)\n","\n","        # print(\"Video out shape : \", video_out.shape)\n","        # print(\"Text input shape : \", text_input.shape)\n","        # weight_a = F.sigmoid(self.acoustic_gate(torch.cat([text_input, audio_out], dim=-1)))\n","        weight_v = F.sigmoid(self.visual_gate(torch.cat([text_input, video_out], dim=-1)))\n","\n","        # output = self.final_layer_norm(text_input + weight_a * audio_out + weight_v * video_out)\n","\n","        output = self.final_layer_norm(text_input  + weight_v * video_out)\n","\n","        return output"]},{"cell_type":"code","source":[],"metadata":{"id":"ctm9L23KByt-","executionInfo":{"status":"ok","timestamp":1684514971232,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":41,"outputs":[]},{"cell_type":"code","execution_count":42,"metadata":{"id":"aIYaqksziiAC","executionInfo":{"status":"ok","timestamp":1684514971232,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MultiModalBartIntentEncoder(BartPretrainedModel):\n","\n","    def __init__(self, config: BartConfig, embed_tokens: Optional[nn.Embedding] = None):\n","        super().__init__(config)\n","\n","        self.dropout = config.dropout\n","        self.layerdrop = config.encoder_layerdrop\n","\n","        embed_dim = config.d_model\n","        self.padding_idx = config.pad_token_id\n","        self.max_source_position = config.max_position_embeddings\n","        self.embed_scale = math.sqrt(embed_dim) if config.scale_embedding else 1.0\n","\n","        if embed_tokens is not None:\n","            self.embed_tokens = embed_tokens\n","        else:\n","            self.embed_tokens = nn.Embedding(config.vocab_size, embed_dim, self.padding_idx)\n","\n","        self.embed_positions = BartLearnedPositionalEmbedding(\n","            config.max_position_embeddings,\n","            embed_dim\n","        )        \n","\n","        \n","        \n","\n","        self.layers = nn.ModuleList([BartEncoderLayer(config) for _ in range(config.encoder_layers)])\n","\n","        self.layernorm_embedding = nn.LayerNorm(embed_dim)\n","\n","        self.init_weights()\n","        self.gradient_checkpointing = False\n","\n","        # self.fusion_at_layer = [4]\n","        # self.fusion_at_layer = [3, 4]\n","        # self.fusion_at_layer3 = [3]\n","        self.fusion_at_layer4 = [4]\n","        self.fusion_at_layer5 = [5]\n","\n","        # self.fusion_of_context = [3]\n","        # self.visual_transformer = TransformerEncoder(d_model = VISUAL_DIM,\n","        #                                              n_layers = 4,\n","        #                                              n_heads=8,\n","        #                                              d_ff=VISUAL_DIM\n","        #                                              )\n","        # self.acoustic_transformer = TransformerEncoder(d_model = ACOUSTIC_DIM,\n","        #                                                n_layers=4,\n","        #                                                n_heads=2,\n","        #                                                d_ff=ACOUSTIC_DIM)\n","\n","        # self.MAF_layer3 = MAF(dim_model=embed_dim,\n","        #                      dropout_rate=0.2)\n","        \n","        self.MAF_layer4 = MAF_Intent_acoustic(dim_model=embed_dim,\n","                             dropout_rate=0.2)\n","        \n","        self.MAF_layer5 = MAF_Intent_visual(dim_model=embed_dim,\n","                             dropout_rate=0.2)\n","        \n","        # self.context_encoder = ContextEncoder(config)\n","\n","        # self.classification = nn.Linear(embed_dim, 2)\n","\n","\n","\n","\n","    def forward(self,\n","            input_ids = None,\n","            attention_mask = None,\n","            # context_input_ids = None,\n","            # context_attention_mask = None,\n","            acoustic_input = None,\n","            visual_input = None,\n","            head_mask = None,\n","            inputs_embeds = None,\n","            output_attentions = None,\n","            output_hidden_states  = None,\n","            return_dict = None):\n","\n","            # print(\"Input ids shape : \", input_ids.shape)\n","            output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions\n","\n","            output_hidden_states = (\n","                output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states\n","            )\n","\n","            return_dict = return_dict if return_dict is not None else self.config.use_return_dict\n","\n","            if input_ids is not None and inputs_embeds is not None:\n","                raise ValueError(\"You can't specify both input_ids and inputs_embeds at the same time\")\n","            elif input_ids is not None:\n","                input_shape = input_ids.size()\n","                input_ids = input_ids.view(-1, input_shape[-1])\n","\n","            elif inputs_embeds is not None:\n","                input_shape = inputs_embeds.size()[:-1]\n","            else:\n","                raise ValueError(\"You have to specify either input_ids or input_embeds\")\n","\n","\n","            if inputs_embeds is None:\n","                # print(\"Input ids shape : \", input_ids.shape)\n","                inputs_embeds = self.embed_tokens(input_ids) * self.embed_scale\n","\n","            # print(\"Input shape type : \", type(input_shape))\n","            # print(\"Input shape : \", input_shape)\n","            input_shape = torch.tensor(input_shape)\n","            # print(\"Input shape type : \", type(input_shape))\n","            # print(\"Input shape : \", input_shape)\n","            # print(\"Input shape : \", input_shape.shape)\n","            embed_pos = self.embed_positions(input_ids)\n","            # embed_pos = self.embed_positions(input_shape)\n","\n","\n","            hidden_states = inputs_embeds + embed_pos   \n","            hidden_states = self.layernorm_embedding(hidden_states)\n","            hidden_states = F.dropout(hidden_states, p = self.dropout, training=self.training)\n","\n","            # print(\"attention mask shape 3 : \", attention_mask.shape)\n","            if attention_mask is not None:\n","                attention_mask =  _expand_mask(attention_mask, inputs_embeds.dtype)\n","\n","            # print(\"attention mask shape 4 : \", attention_mask.shape)\n","            encoder_states = () if output_hidden_states else None\n","            all_attentions = () if output_attentions else None\n","\n","            if head_mask is not None:\n","                assert head_mask.size()[0] == (\n","                    len(self.layers)\n","                ), f\"The head mask should be specified for {len(self.layers)} layers, but it is for {head_mask.size()[0]}.\"\n","\n","            for idx, encoder_layer in enumerate(self.layers):\n","                # print(\"============Idx : \", idx)\n","\n","                # if idx in self.fusion_at_layer3:\n","                #     # print(\"Acoustic input shape (B) : \", acoustic_input)\n","                #     # acoustic_input = self.acoustic_transformer(acoustic_input)[-1]   \n","                #     # print(\"Acoustic input shape (C) : \", acoustic_input)\n","\n","                #     # visual_input = self.visual_transformer(visual_input)[-1]\n","                #     # print(\"====Idx inside fusion at layer :\", idx)\n","                #     hidden_states = self.MAF_layer3(text_input = hidden_states,\n","                #                                    acoustic_context = acoustic_input,\n","                #                                    visual_context = visual_input) \n","\n","                # if idx in self.fusion_of_context:\n","                  \n","                #   hidden_states = self.context_encoder(hidden_states = hidden_states, context_input_ids = context_input_ids, context_attention_mask = context_attention_mask)\n","                  \n","\n","                if idx in self.fusion_at_layer4:\n","                    # print(\"Acoustic input shape (B) : \", acoustic_input)\n","                    # acoustic_input = self.acoustic_transformer(acoustic_input)[-1]   \n","                    # print(\"Acoustic input shape (C) : \", acoustic_input)\n","\n","                    # visual_input = self.visual_transformer(visual_input)[-1]\n","                    # print(\"====Idx inside fusion at layer :\", idx)\n","\n","                    \n","                    \n","                    hidden_states = self.MAF_layer4(text_input = hidden_states,\n","                                                   acoustic_context = acoustic_input\n","                                                   )\n","                if idx in self.fusion_at_layer5:\n","                    # print(\"Acoustic input shape (B) : \", acoustic_input)\n","                    # acoustic_input = self.acoustic_transformer(acoustic_input)[-1]   \n","                    # print(\"Acoustic input shape (C) : \", acoustic_input)\n","\n","                    # visual_input = self.visual_transformer(visual_input)[-1]\n","                    # print(\"====Idx inside fusion at layer :\", idx)\n","                    hidden_states = self.MAF_layer5(text_input = hidden_states,\n","                                                   visual_context = visual_input)     \n","\n","                if output_hidden_states:\n","                    encoder_states = encoder_states + (hidden_states,)\n","\n","                dropout_probability = random.uniform(0,1)   \n","\n","                if self.training and (dropout_probability < self.layerdrop):\n","                    layer_outputs = (None, None)\n","\n","                else:\n","                    if self.gradient_checkpointing and self.training:\n","\n","                        def create_custom_forward(module):\n","                            def custom_forward(*inputs):\n","                                return module(*inputs, output_attentions)\n","\n","                            return custom_forward\n","\n","                        layer_outputs = torch.utils.checkpoint.checkpoint(\n","                            create_custom_forward(encoder_layer),\n","                            hidden_states,\n","                            attention_mask,\n","                            (head_mask[idx] if head_mask is not None else None),\n","                        )        \n","\n","                    else:\n","                        # print(\"Checking Attention mask shape : \", attention_mask.shape)\n","                        layer_outputs = encoder_layer(\n","                            hidden_states,\n","                            attention_mask,\n","                            layer_head_mask = (head_mask[idx] if head_mask is not None else None),\n","                            output_attentions = output_attentions\n","                        )    \n","\n","                    hidden_states = layer_outputs[0]\n","\n","                if output_attentions:\n","                    all_attentions  = all_attentions + (layer_outputs[1],)\n","\n","            if output_hidden_states:\n","                encoder_states = encoder_states + (hidden_states,)\n","\n","            if not return_dict:\n","                return tuple(v for v in [hidden_states, encoder_states, all_attentions] if v is not None)\n","\n","            return BaseModelOutput(\n","                last_hidden_state=hidden_states, hidden_states=encoder_states, attentions=all_attentions\n","            )    \n","\n","            # print(\"Hidden states shape : \", hidden_states)\n","\n","            # cls = hidden_states.permute(1,0,2)"]},{"cell_type":"code","execution_count":43,"metadata":{"id":"KLseqaTKijxS","executionInfo":{"status":"ok","timestamp":1684514971232,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MultimodalBartIntentModel(BartPretrainedModel):\n","    def __init__(self, config: BartConfig):\n","        super().__init__(config)\n","\n","        padding_idx, vocab_size = config.pad_token_id, config.vocab_size\n","        self.shared = nn.Embedding(vocab_size, config.d_model, padding_idx)\n","\n","        self.encoder = MultiModalBartIntentEncoder(config, self.shared)\n","        self.decoder = BartDecoder(config, self.shared)\n","\n","        self.init_weights()\n","\n","    def get_input_embeddings(self):\n","        return self.shared\n","\n","    def set_input_embeddings(self, value):\n","        self.shared = value\n","        self.encoder.embed_tokens = self.shared\n","        self.decoder.embed_tokens = self.shared\n","\n","    def get_encoder(self):\n","        return self.encoder\n","\n","\n","    def get_decoder(self):\n","        return self.decoder\n","\n","    def forward(\n","        self, \n","        input_ids = None,\n","        attention_mask = None,\n","        # context_input_ids = None,\n","        # context_attention_mask = None,\n","        acoustic_input = None,\n","        visual_input = None,\n","        decoder_input_ids = None,\n","        decoder_attention_mask = None,\n","        head_mask = None,\n","        decoder_head_mask = None,\n","        cross_attn_head_mask = None,\n","        encoder_outputs = None,\n","        past_key_values = None,\n","        inputs_embeds = None,\n","        decoder_inputs_embeds = None,\n","        use_cache = None,\n","        output_attentions = None,\n","        output_hidden_states = None,\n","        return_dict = None\n","    ):\n","\n","        if decoder_input_ids is None and decoder_inputs_embeds is None:\n","            decoder_input_ids = shift_tokens_right(\n","                input_ids, self.config.pad_token_id, self.config.decoder_start_token_id\n","\n","            )                                \n","\n","        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions\n","        output_hidden_states = (\n","            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states\n","        )    \n","        use_cache = use_cache if use_cache is not None else self.config.use_cache\n","        return_dict = return_dict if return_dict is not None else self.config.use_return_dict\n","\n","        # print(\"attention mask shape 2 : \", attention_mask.shape)\n","\n","        if encoder_outputs is None:\n","            encoder_outputs = self.encoder(\n","                input_ids = input_ids,\n","                attention_mask = attention_mask,\n","                # context_input_ids = context_input_ids,\n","                # context_attention_mask = context_attention_mask,\n","                acoustic_input = acoustic_input,\n","                visual_input = visual_input,\n","                head_mask = head_mask,\n","                inputs_embeds = inputs_embeds,\n","                output_attentions = output_attentions,\n","                output_hidden_states = output_hidden_states,\n","                return_dict = return_dict\n","            )\n","\n","        elif return_dict and not isinstance(encoder_outputs, BaseModelOutput):\n","            encoder_outputs = BaseModelOutput(\n","                last_hidden_state=encoder_outputs[0],\n","                hidden_states=encoder_outputs[1] if len(encoder_outputs) > 1 else None,\n","                attentions = encoder_outputs[2] if len(encoder_outputs) > 2 else None\n","            )    \n","\n","        decoder_outputs = self.decoder(\n","            input_ids = decoder_input_ids,\n","            attention_mask = decoder_attention_mask,\n","            encoder_hidden_states = encoder_outputs[0],\n","            encoder_attention_mask = attention_mask,\n","            head_mask = decoder_head_mask,\n","            cross_attn_head_mask = cross_attn_head_mask,\n","            past_key_values = past_key_values,\n","            inputs_embeds = decoder_inputs_embeds,\n","            use_cache = use_cache,\n","            output_attentions = output_attentions,\n","            output_hidden_states = output_hidden_states,\n","            return_dict = return_dict\n","        )    \n","\n","        if not return_dict:\n","            return decoder_outputs + encoder_outputs\n","\n","        return Seq2SeqModelOutput(\n","            last_hidden_state=decoder_outputs.last_hidden_state,\n","            past_key_values=decoder_outputs.past_key_values,\n","            decoder_hidden_states=decoder_outputs.hidden_states,\n","            decoder_attentions=decoder_outputs.attentions,\n","            cross_attentions=decoder_outputs.cross_attentions,\n","            encoder_last_hidden_state=encoder_outputs.last_hidden_state,\n","            encoder_hidden_states=encoder_outputs.hidden_states,\n","            encoder_attentions=encoder_outputs.attentions\n","\n","        )"]},{"cell_type":"code","execution_count":44,"metadata":{"id":"wTixVwZZilVS","executionInfo":{"status":"ok","timestamp":1684514971232,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MultimodalBartIntentClassification(nn.Module):\n","    def __init__(\n","        self, \n","        input_dim: int,\n","        inned_dim: int,\n","        num_classes: int,\n","        pooler_dropout: float\n","    ):\n","        super().__init__()\n","        self.dense = nn.Linear(input_dim, inned_dim)\n","        self.dropout = nn.Dropout(p = pooler_dropout)\n","        self.out_proj = nn.Linear(inned_dim, num_classes)\n","\n","    def forward(self, hidden_states):\n","        hidden_states = self.dropout(hidden_states)\n","        hidden_states = self.dense(hidden_states)\n","        hidden_states = torch.tanh(hidden_states)\n","        temp_hid = hidden_states\n","        hidden_states = self.dropout(hidden_states)\n","        hidden_states = self.out_proj(hidden_states)\n","        return hidden_states, temp_hid    \n","\n","\n","\n","\n","class MultimodalBartIntentForSequenceClassification(BartPretrainedModel):\n","    _keys_to_ignore_on_load_missing = [\"encoder.embed_tokens.weight\", \"decoder.embed_tokens.weight\"]\n","\n","    def __init__(self, config: BartConfig, **kwargs):\n","        super().__init__(config, **kwargs)\n","        self.model = MultimodalBartIntentModel(config)\n","        self.classification_head = MultimodalBartIntentClassification(\n","            config.d_model,\n","            config.d_model,\n","            12,\n","            config.classifier_dropout\n","        )\n","        self.model._init_weights(self.classification_head.dense)\n","        self.model._init_weights(self.classification_head.out_proj)\n","\n","    def forward(\n","        self,\n","        input_ids: torch.LongTensor = None,\n","        attention_mask: Optional[torch.tensor] = None,\n","        # context_input_ids : torch.LongTensor = None,\n","        # context_attention_mask : Optional[torch.tensor] = None,\n","        acoustic_input = None,\n","        visual_input = None,\n","        decoder_input_ids: Optional[torch.LongTensor] = None,\n","        decoder_attention_mask: Optional[torch.LongTensor] = None,\n","        head_mask: Optional[torch.Tensor] = None,\n","        decoder_head_mask: Optional[torch.Tensor] = None,\n","        cross_attn_head_mask: Optional[torch.Tensor] = None,\n","        encoder_outputs: Optional[List[torch.FloatTensor]] = None,\n","        inputs_embeds: Optional[torch.FloatTensor] = None,\n","        decoder_inputs_embeds: Optional[torch.FloatTensor] = None,\n","        labels: Optional[torch.LongTensor] = None,\n","        use_cache: Optional[bool] = None,\n","        output_attentions: Optional[bool] = None,\n","        output_hidden_states: Optional[bool] = None,\n","        return_dict: Optional[bool] = None\n","    ) -> Union[Tuple, Seq2SeqSequenceClassifierOutput]:\n","        r\"\"\"\n","        labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):\n","            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,\n","            config.num_labels - 1]`. If `config.num_labels > 1` a classification loss is computed (Cross-Entropy).\n","        \"\"\"\n","\n","        return_dict = return_dict if return_dict is not None else self.config.use_return_dict\n","        if labels is not None:\n","            use_cache = False\n","\n","        if input_ids is None and inputs_embeds is not None:\n","            raise NotImplementedError(\n","                f\"Passing input embeddings is currently not supported for {self.__class__.__name__}\"\n","            )\n","\n","        # print(\"attention mask shape 1 : \", attention_mask.shape )\n","        outputs = self.model(\n","            input_ids,\n","            attention_mask = attention_mask,\n","            # context_input_ids = context_input_ids,\n","            # context_attention_mask = context_attention_mask,\n","            acoustic_input = acoustic_input,\n","            visual_input = visual_input,\n","            decoder_input_ids = decoder_input_ids,\n","            decoder_attention_mask = decoder_attention_mask,\n","            head_mask = head_mask,\n","            decoder_head_mask = decoder_head_mask,\n","            cross_attn_head_mask = cross_attn_head_mask,\n","            encoder_outputs = encoder_outputs,\n","            inputs_embeds = inputs_embeds,\n","            decoder_inputs_embeds = decoder_inputs_embeds,\n","            use_cache = use_cache,\n","            output_attentions = output_attentions,\n","            output_hidden_states = output_hidden_states,\n","            return_dict = return_dict\n","\n","        )    \n","\n","        hidden_states = outputs[0]\n","\n","        eos_mask = input_ids.eq(self.config.eos_token_id).to(hidden_states.device)\n","\n","        if len(torch.unique_consecutive(eos_mask.sum(1))) > 1:\n","            raise ValueError(\"All examples must have same number of <eos> tokens\")\n","\n","        sentence_representation = hidden_states[eos_mask, :].view(hidden_states.size(0), -1, hidden_states.size(-1))[\n","            :, -1, :\n","        ]    \n","\n","        logits, temp_hid = self.classification_head(sentence_representation)\n","\n","        loss = None\n","\n","        loss_fct = CrossEntropyLoss()\n","        # print(\"logits shape : \", logits.shape)\n","        loss = loss_fct(logits.view(-1, 12), labels.view(-1))\n","        # print(\"Loss : \", loss)\n","        if not return_dict:\n","            output = (logits,) + outputs[1:]\n","            return ((loss,) + output) if loss is not None else output\n","\n","\n","        return Seq2SeqSequenceClassifierOutput(\n","            loss=loss,\n","            logits=logits,\n","            past_key_values=outputs.past_key_values,\n","            decoder_hidden_states=outputs.decoder_hidden_states,\n","            decoder_attentions = outputs.decoder_attentions,\n","            cross_attentions=outputs.cross_attentions,\n","            encoder_last_hidden_state=outputs.encoder_last_hidden_state,\n","            encoder_hidden_states=outputs.encoder_hidden_states,\n","            encoder_attentions=outputs.encoder_attentions\n","        ), sentence_representation, hidden_states, temp_hid "]},{"cell_type":"code","execution_count":45,"metadata":{"id":"m4lcNczUHpOH","executionInfo":{"status":"ok","timestamp":1684514971233,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# def audio_video_broadcast(x):\n","# #     z = torch.empty()\n","#     temp_all = torch.Tensor()\n","#     for j in range(x.shape[0]):\n","#         print(\"j : \", j)\n","#         temp_x = x[j,:]\n","# #         print(\"Temp x shape : \", temp_x.shape)\n","#         temp_x = torch.tensor(temp_x, dtype=torch.float)\n","#         temp_x = torch.broadcast_to(temp_x, (SOURCE_MAX_LEN, temp_x.shape[0]))\n","# #         print(\"Temp x shape : \", temp_x.shape)\n","#         temp_x = temp_x.unsqueeze(0)\n","# #         print(\"Temp x shape : \", temp_x.shape)\n","        \n","#         if(j==0):\n","#             temp_all = temp_x\n","#         else:\n","#             temp_all = torch.cat([temp_all, temp_x], dim = 0)\n","        \n","        \n","#     return temp_all "]},{"cell_type":"code","execution_count":45,"metadata":{"id":"inWKX8K6GV50","executionInfo":{"status":"ok","timestamp":1684514971233,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":[]},{"cell_type":"code","execution_count":46,"metadata":{"id":"KaewwlapryS5","executionInfo":{"status":"ok","timestamp":1684514971233,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# with open('/content/drive/MyDrive/Colab Notebooks/32/train_dlog_id.p', 'rb') as f:\n","#   train_dlog_id = pickle.load(f)"]},{"cell_type":"code","execution_count":47,"metadata":{"id":"P9etiInqr4-9","executionInfo":{"status":"ok","timestamp":1684514971233,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# with open('/content/drive/MyDrive/Colab Notebooks/32/test_dlog_id.p', 'rb') as f:\n","#   test_dlog_id = pickle.load(f)"]},{"cell_type":"code","execution_count":48,"metadata":{"id":"RunFr5b0Ni1u","executionInfo":{"status":"ok","timestamp":1684514971233,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# with open('/content/drive/MyDrive/Colab Notebooks/32/train_audio_fold_'+str(foldNum)+'.p', 'rb') as f:\n","#   train_audio_data_utterance1 = pickle.load(f)\n"]},{"cell_type":"code","execution_count":49,"metadata":{"id":"pGFk4HCwOo-A","executionInfo":{"status":"ok","timestamp":1684514971234,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# type(train_audio_data_utterance1)"]},{"cell_type":"code","execution_count":50,"metadata":{"id":"sK4GbT8ZOq2W","executionInfo":{"status":"ok","timestamp":1684514971234,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# len(train_audio_data_utterance1)"]},{"cell_type":"code","execution_count":50,"metadata":{"id":"X73Hp5HxOsEl","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":[]},{"cell_type":"code","execution_count":50,"metadata":{"id":"yM1zXLalOifj","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":[]},{"cell_type":"code","execution_count":51,"metadata":{"id":"tvBQbRyFN5zu","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# with open('/content/drive/MyDrive/Colab Notebooks/32/test_audio_fold_'+str(foldNum)+'.p', 'rb') as f:\n","#   test_audio_data_utterance1 = pickle.load(f)"]},{"cell_type":"code","execution_count":52,"metadata":{"id":"ZctmFGmNQsMN","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# len(test_audio_data_utterance1)"]},{"cell_type":"code","execution_count":53,"metadata":{"id":"Dz2xdgV7PzbN","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# len(test_audio_data_utterance1)"]},{"cell_type":"code","execution_count":54,"metadata":{"id":"Yw0sc0aaNzOO","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# with open('/content/drive/MyDrive/Colab Notebooks/32/train_video_fold_'+str(foldNum)+'.p', 'rb') as f:\n","#   train_image_data_utterance1 = pickle.load(f)"]},{"cell_type":"code","execution_count":55,"metadata":{"id":"0yT8QM77QvJS","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# len(train_image_data_utterance1)"]},{"cell_type":"code","execution_count":56,"metadata":{"id":"aUEvmMBNOPFE","executionInfo":{"status":"ok","timestamp":1684514971235,"user_tz":-330,"elapsed":37,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# with open('/content/drive/MyDrive/Colab Notebooks/32/test_video_fold_'+str(foldNum)+'.p', 'rb') as f:\n","#   test_image_data_utterance1 = pickle.load(f)"]},{"cell_type":"code","execution_count":57,"metadata":{"id":"gm6Z6Fc0QxSW","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":38,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# len(test_image_data_utterance1)"]},{"cell_type":"code","execution_count":58,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":38,"status":"ok","timestamp":1684514971236,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"KIEMeHJ6RoXS","outputId":"decbd21f-f7bb-460b-e614-bdc3387c240e"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["tensor([[1., 1., 1., 1., 1.],\n","        [1., 1., 1., 1., 1.],\n","        [1., 1., 1., 1., 1.],\n","        [1., 1., 1., 1., 1.]])"]},"metadata":{},"execution_count":58}],"source":["tp = torch.ones(4,5)\n","tp"]},{"cell_type":"code","execution_count":59,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":34,"status":"ok","timestamp":1684514971236,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"o8vsmq4FS48j","outputId":"9ea0edcd-8cfc-4d38-8567-eab97ee53d1b"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["tensor([[0., 0., 0., 0., 0.]])"]},"metadata":{},"execution_count":59}],"source":["torch.zeros(5 - tp.shape[0], 5)"]},{"cell_type":"code","execution_count":60,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":33,"status":"ok","timestamp":1684514971236,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"6CNdKCeQRrN8","outputId":"03d52fd4-91ce-43ab-9887-339e6040e3f0"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["tensor([[1., 1., 1., 1., 1.],\n","        [1., 1., 1., 1., 1.],\n","        [1., 1., 1., 1., 1.],\n","        [1., 1., 1., 1., 1.],\n","        [0., 0., 0., 0., 0.]])"]},"metadata":{},"execution_count":60}],"source":["torch.cat([tp, torch.zeros(5 - tp.shape[0], 5)])"]},{"cell_type":"code","execution_count":61,"metadata":{"id":"qO9elV8EQAi1","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# def pad_seq(tensor, dim, max_len):\n","#   if max_len > tensor.shape[0] :\n","#     return torch.cat([tensor, torch.zeros(max_len - tensor.shape[0], dim)])\n","#   else:\n","#     return tensor[:max_len]  "]},{"cell_type":"code","execution_count":62,"metadata":{"id":"HmBt6xZ1TElw","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# ACOUSTIC_DIM"]},{"cell_type":"code","execution_count":63,"metadata":{"id":"4cUPe2LxQ04S","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# train_audio_data_utterance1 = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","#                                                    dim = ACOUSTIC_DIM,\n","#                                                    max_len = ACOUSTIC_MAX_LEN)\n","#                                                   for a in train_audio_data_utterance1], 0)\n","# train_audio_data_utterance1.shape"]},{"cell_type":"code","execution_count":64,"metadata":{"id":"x6-frFIpTzrW","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# test_audio_data_utterance1 = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","#                                                    dim = ACOUSTIC_DIM,\n","#                                                    max_len = ACOUSTIC_MAX_LEN)\n","#                                                   for a in test_audio_data_utterance1], 0)\n","# test_audio_data_utterance1.shape"]},{"cell_type":"code","execution_count":65,"metadata":{"id":"WEDVn1TaUicC","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# VISUAL_DIM"]},{"cell_type":"code","execution_count":66,"metadata":{"id":"7ArX7k9VUkWO","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":30,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# VISUAL_MAX_LEN"]},{"cell_type":"code","execution_count":67,"metadata":{"id":"PO2OLh5yUTF2","executionInfo":{"status":"ok","timestamp":1684514971236,"user_tz":-330,"elapsed":30,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# train_image_data_utterance1 = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","#                                                    dim = VISUAL_DIM,\n","#                                                    max_len = VISUAL_MAX_LEN)\n","#                                                   for a in train_image_data_utterance1], 0)\n","# train_image_data_utterance1.shape"]},{"cell_type":"code","execution_count":68,"metadata":{"id":"9BAAQoCcUv34","executionInfo":{"status":"ok","timestamp":1684514971237,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# test_image_data_utterance1 = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","#                                                    dim = VISUAL_DIM,\n","#                                                    max_len = VISUAL_MAX_LEN)\n","#                                                   for a in test_image_data_utterance1], 0)\n","# test_image_data_utterance1.shape"]},{"cell_type":"code","execution_count":68,"metadata":{"id":"TTMNyPhSUdBs","executionInfo":{"status":"ok","timestamp":1684514971237,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":[]},{"cell_type":"code","execution_count":69,"metadata":{"id":"j3V8UO6HH0xE","executionInfo":{"status":"ok","timestamp":1684514971237,"user_tz":-330,"elapsed":31,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# path = \"/content/drive/MyDrive/Colab Notebooks/32/datasetTrue_original/sarcasmDataset_speaker_dependent_True.npz\"\n","# data2 = np.load(path, mmap_mode=True)   \n","\n","# train_audio_data_utterance1 = data2['feautesUA_train'][foldNum]\n","# train_image_data_utterance1 = data2['feautesUV_train'][foldNum]\n","\n","# test_audio_data_utterance1 = data2['feautesUA_test'][foldNum]\n","# test_image_data_utterance1 = data2['feautesUV_test'][foldNum]"]},{"cell_type":"code","execution_count":70,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000,"referenced_widgets":["8ebacbe3dc344455befd54fc5665fd1b","3133a56799144a108de38093072999cd","a7e0f241ef744f4bae89846070aac64b","0ae280d726d444bbae665c7fb4994a29","a928128d4e364b6dbc40e45acdc5fdd5","2004beaac327433a88c0cc186b6da8be","00d79bb4d88e45ce8a2361ca3eda44bb","58085460ac1e479581252e38ff3b5340","93db01d807dd4250b3c14d3746ce14a2","3b3ddba2572c42b5ab5e6da9e19e9570","89b3c334d2cf4b89b23932456c012a54","dfdc980705b746898f6db4a14086b5df","7e4680003db447f6972e46f8e9bf3ce4","bb530047087042c79a8af0b97c6c19ed","012dacacfdcf4ec1b8b13106d4e65bae","17066910932d4524a4b8fbb25e0a54e2","30f1f20b18524603923227792608a355","4d24131933004629a69c0fd75bcb7ec3","0e4f53830ba9473eaa7346c253d24c06","ab656a90064e41889c53c640acb00e4a","625701156ac74b7fa66a5dd857495ed0","69214f05eee74a9dacf3bf3dd42ed458","e266f9a9eb5b481f9be0b0f407073db6","e06b31e87af449ed90dc814b2d5168a4","c89ca5e444654d0ba11faf028a06a57f","796e8c3bee6b4da294ca53c4253fc189","1450831a48954df4a26a7d3248e2a9bd","c301f9b100ad4d2f902072a7100afadb","c1ea5bbf73bf4405898bd93120e40a41","d7833fcf43b345fd8b31de3d4d4c640f","ca554554f6594d17b4470c270f8f894a","6b32ead6e79a433d9f5355efd2bf68e6","b3e3079a882f43c9adcb9c1e9b51997c","cb2ae3714da14a86a6fac3931bc5fa8b","f20d3416e5b34a3ea1b574dcc3e242e6","f52a8877833f4157b873d07aa36c057d","0b24ad5aa31f4ee5ab95a8476397af1c","ef45571d08a24746afaca14cd1a17a4e","27437b8101444ef5a355450c67e6af25","dc64c16ebc9a465b8d847a522f70cdb2","d6e9277a98e74805ad6cd54936e18f77","02a9ae0e473d4ff59c897d41fa03dcda","083c3a66fba34396bd1a0e50e4bed84d","6eee378f1c7b45d0b9cc79b5776ddaea","391921869d734be5adc96f25de92e793","dbb378ddfc914fa98b47e4d0e3a37be6","5fb909faa6924c0ea137eb9d5419144b","d79c10248bb94161bdbdf15f9e76fe76","3da0e0b652224cbb9414afbab0c235ba","0a555dcee82046898f35ce03e4fb7ab4","5164a93bf0ac411782e2ca32d4a9742a","0b7fbae40e5a4a299ab28540c3a4083f","a4a111be2f264f7992f28427270ed9b5","28c2cab9b929414496329971a1e564bc","8591e8ab909641098c641ef244bbbce2"]},"executionInfo":{"elapsed":13128,"status":"ok","timestamp":1684514984334,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"pdf5hGlVH5Zw","outputId":"eca15442-a2f8-4dcd-a373-d3310fd33b4a"},"outputs":[{"output_type":"display_data","data":{"text/plain":["Downloading (…)lve/main/config.json:   0%|          | 0.00/1.72k [00:00<?, ?B/s]"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"8ebacbe3dc344455befd54fc5665fd1b"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["Downloading pytorch_model.bin:   0%|          | 0.00/558M [00:00<?, ?B/s]"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"dfdc980705b746898f6db4a14086b5df"}},"metadata":{}},{"output_type":"stream","name":"stderr","text":["Some weights of MultimodalBartForSequenceClassification were not initialized from the model checkpoint at facebook/bart-base and are newly initialized: ['classification_head.dense.bias', 'encoder.MAF_layer4.acoustic_context_attention.w1_v.weight', 'encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_weight', 'encoder.MAF_layer5.visual_context_attention.u_v.weight', 'encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.weight', 'encoder.MAF_layer5.visual_context_attention.w2_k.weight', 'encoder.MAF_layer5.visual_gate.weight', 'encoder.MAF_layer4.acoustic_context_transform.weight', 'encoder.MAF_layer4.acoustic_context_attention.u_k.weight', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_bias', 'encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_bias', 'encoder.MAF_layer5.visual_context_attention.w2_v.weight', 'encoder.MAF_layer5.visual_gate.bias', 'encoder.MAF_layer4.acoustic_gate.bias', 'classification_head.out_proj.weight', 'encoder.MAF_layer5.visual_context_attention.w1_v.weight', 'encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.bias', 'encoder.MAF_layer5.visual_context_attention.u_k.weight', 'encoder.MAF_layer4.final_layer_norm.weight', 'encoder.MAF_layer4.acoustic_context_attention.w2_k.weight', 'classification_head.dense.weight', 'encoder.MAF_layer4.acoustic_gate.weight', 'classification_head.out_proj.bias', 'encoder.MAF_layer4.acoustic_context_attention.w2_v.weight', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.bias', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_weight', 'encoder.MAF_layer4.final_layer_norm.bias', 'encoder.MAF_layer5.final_layer_norm.weight', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.weight', 'encoder.MAF_layer5.final_layer_norm.bias', 'encoder.MAF_layer4.acoustic_context_attention.w1_k.weight', 'encoder.MAF_layer4.acoustic_context_attention.u_v.weight', 'encoder.MAF_layer5.visual_context_transform.weight', 'encoder.MAF_layer5.visual_context_attention.w1_k.weight']\n","You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"]},{"output_type":"stream","name":"stdout","text":["MultimodalBartForSequenceClassification(\n","  (model): MultimodalBartModel(\n","    (shared): Embedding(50265, 768, padding_idx=1)\n","    (encoder): MultiModalBartEncoder(\n","      (embed_tokens): Embedding(50265, 768, padding_idx=1)\n","      (embed_positions): BartLearnedPositionalEmbedding(1026, 768)\n","      (layers): ModuleList(\n","        (0-5): 6 x BartEncoderLayer(\n","          (self_attn): BartAttention(\n","            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n","          )\n","          (self_attn_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","          (activation_fn): GELUActivation()\n","          (fc1): Linear(in_features=768, out_features=3072, bias=True)\n","          (fc2): Linear(in_features=3072, out_features=768, bias=True)\n","          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","        )\n","      )\n","      (layernorm_embedding): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","      (MAF_layer4): MAF_acoustic(\n","        (acoustic_context_transform): Linear(in_features=1000, out_features=234, bias=False)\n","        (acoustic_context_attention): ContextAwareAttention(\n","          (attention_layer): MultiheadAttention(\n","            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n","          )\n","          (u_k): Linear(in_features=768, out_features=768, bias=False)\n","          (w1_k): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_k): Linear(in_features=768, out_features=1, bias=False)\n","          (u_v): Linear(in_features=768, out_features=768, bias=False)\n","          (w1_v): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_v): Linear(in_features=768, out_features=1, bias=False)\n","        )\n","        (acoustic_gate): Linear(in_features=1536, out_features=768, bias=True)\n","        (dropout_layer): Dropout(p=0.2, inplace=False)\n","        (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","      )\n","      (MAF_layer5): MAF_visual(\n","        (visual_context_transform): Linear(in_features=480, out_features=234, bias=False)\n","        (visual_context_attention): ContextAwareAttention(\n","          (attention_layer): MultiheadAttention(\n","            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n","          )\n","          (u_k): Linear(in_features=2048, out_features=768, bias=False)\n","          (w1_k): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_k): Linear(in_features=768, out_features=1, bias=False)\n","          (u_v): Linear(in_features=2048, out_features=768, bias=False)\n","          (w1_v): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_v): Linear(in_features=768, out_features=1, bias=False)\n","        )\n","        (visual_gate): Linear(in_features=1536, out_features=768, bias=True)\n","        (dropout_layer): Dropout(p=0.2, inplace=False)\n","        (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","      )\n","    )\n","    (decoder): BartDecoder(\n","      (embed_tokens): Embedding(50265, 768, padding_idx=1)\n","      (embed_positions): BartLearnedPositionalEmbedding(1026, 768)\n","      (layers): ModuleList(\n","        (0-5): 6 x BartDecoderLayer(\n","          (self_attn): BartAttention(\n","            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n","          )\n","          (activation_fn): GELUActivation()\n","          (self_attn_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","          (encoder_attn): BartAttention(\n","            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n","          )\n","          (encoder_attn_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","          (fc1): Linear(in_features=768, out_features=3072, bias=True)\n","          (fc2): Linear(in_features=3072, out_features=768, bias=True)\n","          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","        )\n","      )\n","      (layernorm_embedding): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","    )\n","  )\n","  (classification_head): MultimodalBartClassification(\n","    (dense): Linear(in_features=768, out_features=768, bias=True)\n","    (dropout): Dropout(p=0.0, inplace=False)\n","    (out_proj): Linear(in_features=768, out_features=2, bias=True)\n","  )\n",")\n"]},{"output_type":"display_data","data":{"text/plain":["Downloading (…)olve/main/vocab.json:   0%|          | 0.00/899k [00:00<?, ?B/s]"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"e266f9a9eb5b481f9be0b0f407073db6"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["Downloading (…)olve/main/merges.txt:   0%|          | 0.00/456k [00:00<?, ?B/s]"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"cb2ae3714da14a86a6fac3931bc5fa8b"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["Downloading (…)/main/tokenizer.json:   0%|          | 0.00/1.36M [00:00<?, ?B/s]"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"391921869d734be5adc96f25de92e793"}},"metadata":{}},{"output_type":"stream","name":"stdout","text":["Tokenizer :  BartTokenizerFast(name_or_path='facebook/bart-base', vocab_size=50265, model_max_length=1024, is_fast=True, padding_side='right', truncation_side='right', special_tokens={'bos_token': '<s>', 'eos_token': '</s>', 'unk_token': '<unk>', 'sep_token': '</s>', 'pad_token': '<pad>', 'cls_token': '<s>', 'mask_token': AddedToken(\"<mask>\", rstrip=False, lstrip=True, single_word=False, normalized=False)}, clean_up_tokenization_spaces=True)\n","Total parameters :  151.779026\n"]}],"source":["model_sarcasm_only = MultimodalBartForSequenceClassification.from_pretrained(\"facebook/bart-base\")\n","print(model_sarcasm_only)\n","\n","tokenizer = BartTokenizerFast.from_pretrained('facebook/bart-base')\n","print(\"Tokenizer : \", tokenizer)\n","\n","num_param = sum(p.numel() for p in model_sarcasm_only.parameters())\n","print(\"Total parameters : \", num_param/1e6)"]},{"cell_type":"code","execution_count":71,"metadata":{"id":"XRc6ENbdiR6y","executionInfo":{"status":"ok","timestamp":1684514984334,"user_tz":-330,"elapsed":7,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# model_sarcasm_shared = SarcasmShared()\n","model_intent_shared = IntentShared()"]},{"cell_type":"code","execution_count":72,"metadata":{"id":"6IDC7AUNjuCn","executionInfo":{"status":"ok","timestamp":1684514984334,"user_tz":-330,"elapsed":6,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# model_sarcasm_shared"]},{"cell_type":"code","execution_count":73,"metadata":{"executionInfo":{"elapsed":6,"status":"ok","timestamp":1684514984334,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"wx_DUgcgjwON","colab":{"base_uri":"https://localhost:8080/"},"outputId":"d8d844ed-7542-4999-8f31-29eda5d416dc"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["IntentShared(\n","  (gate): Linear(in_features=1536, out_features=768, bias=True)\n","  (classifier): Linear(in_features=768, out_features=12, bias=True)\n",")"]},"metadata":{},"execution_count":73}],"source":["model_intent_shared"]},{"cell_type":"code","execution_count":74,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":2444,"status":"ok","timestamp":1684514986774,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"5cr6qMTfjxnj","outputId":"9dbf4326-f180-4817-fe95-a7cf0bb232c2"},"outputs":[{"output_type":"stream","name":"stderr","text":["Some weights of MultimodalBartIntentForSequenceClassification were not initialized from the model checkpoint at facebook/bart-base and are newly initialized: ['classification_head.dense.bias', 'encoder.MAF_layer4.acoustic_context_attention.w1_v.weight', 'encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_weight', 'encoder.MAF_layer5.visual_context_attention.u_v.weight', 'encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.weight', 'encoder.MAF_layer5.visual_context_attention.w2_k.weight', 'encoder.MAF_layer5.visual_gate.weight', 'encoder.MAF_layer4.acoustic_context_transform.weight', 'encoder.MAF_layer4.acoustic_context_attention.u_k.weight', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_bias', 'encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_bias', 'encoder.MAF_layer5.visual_context_attention.w2_v.weight', 'encoder.MAF_layer5.visual_gate.bias', 'encoder.MAF_layer4.acoustic_gate.bias', 'classification_head.out_proj.weight', 'encoder.MAF_layer5.visual_context_attention.w1_v.weight', 'encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.bias', 'encoder.MAF_layer5.visual_context_attention.u_k.weight', 'encoder.MAF_layer4.final_layer_norm.weight', 'encoder.MAF_layer4.acoustic_context_attention.w2_k.weight', 'classification_head.dense.weight', 'encoder.MAF_layer4.acoustic_gate.weight', 'classification_head.out_proj.bias', 'encoder.MAF_layer4.acoustic_context_attention.w2_v.weight', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.bias', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_weight', 'encoder.MAF_layer4.final_layer_norm.bias', 'encoder.MAF_layer5.final_layer_norm.weight', 'encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.weight', 'encoder.MAF_layer5.final_layer_norm.bias', 'encoder.MAF_layer4.acoustic_context_attention.w1_k.weight', 'encoder.MAF_layer4.acoustic_context_attention.u_v.weight', 'encoder.MAF_layer5.visual_context_transform.weight', 'encoder.MAF_layer5.visual_context_attention.w1_k.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_intent_only = MultimodalBartIntentForSequenceClassification.from_pretrained('facebook/bart-base')"]},{"cell_type":"code","execution_count":75,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":36,"status":"ok","timestamp":1684514986775,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"ad91XS__82SC","outputId":"8e5a595a-9f81-422a-da4c-93a0027996a1"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["MultimodalBartIntentForSequenceClassification(\n","  (model): MultimodalBartIntentModel(\n","    (shared): Embedding(50265, 768, padding_idx=1)\n","    (encoder): MultiModalBartIntentEncoder(\n","      (embed_tokens): Embedding(50265, 768, padding_idx=1)\n","      (embed_positions): BartLearnedPositionalEmbedding(1026, 768)\n","      (layers): ModuleList(\n","        (0-5): 6 x BartEncoderLayer(\n","          (self_attn): BartAttention(\n","            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n","          )\n","          (self_attn_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","          (activation_fn): GELUActivation()\n","          (fc1): Linear(in_features=768, out_features=3072, bias=True)\n","          (fc2): Linear(in_features=3072, out_features=768, bias=True)\n","          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","        )\n","      )\n","      (layernorm_embedding): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","      (MAF_layer4): MAF_Intent_acoustic(\n","        (acoustic_context_transform): Linear(in_features=1000, out_features=234, bias=False)\n","        (acoustic_context_attention): IntentContextAwareAttention(\n","          (attention_layer): MultiheadAttention(\n","            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n","          )\n","          (u_k): Linear(in_features=768, out_features=768, bias=False)\n","          (w1_k): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_k): Linear(in_features=768, out_features=1, bias=False)\n","          (u_v): Linear(in_features=768, out_features=768, bias=False)\n","          (w1_v): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_v): Linear(in_features=768, out_features=1, bias=False)\n","        )\n","        (acoustic_gate): Linear(in_features=1536, out_features=768, bias=True)\n","        (dropout_layer): Dropout(p=0.2, inplace=False)\n","        (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","      )\n","      (MAF_layer5): MAF_Intent_visual(\n","        (visual_context_transform): Linear(in_features=480, out_features=234, bias=False)\n","        (visual_context_attention): IntentContextAwareAttention(\n","          (attention_layer): MultiheadAttention(\n","            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n","          )\n","          (u_k): Linear(in_features=2048, out_features=768, bias=False)\n","          (w1_k): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_k): Linear(in_features=768, out_features=1, bias=False)\n","          (u_v): Linear(in_features=2048, out_features=768, bias=False)\n","          (w1_v): Linear(in_features=768, out_features=1, bias=False)\n","          (w2_v): Linear(in_features=768, out_features=1, bias=False)\n","        )\n","        (visual_gate): Linear(in_features=1536, out_features=768, bias=True)\n","        (dropout_layer): Dropout(p=0.2, inplace=False)\n","        (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","      )\n","    )\n","    (decoder): BartDecoder(\n","      (embed_tokens): Embedding(50265, 768, padding_idx=1)\n","      (embed_positions): BartLearnedPositionalEmbedding(1026, 768)\n","      (layers): ModuleList(\n","        (0-5): 6 x BartDecoderLayer(\n","          (self_attn): BartAttention(\n","            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n","          )\n","          (activation_fn): GELUActivation()\n","          (self_attn_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","          (encoder_attn): BartAttention(\n","            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n","            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n","          )\n","          (encoder_attn_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","          (fc1): Linear(in_features=768, out_features=3072, bias=True)\n","          (fc2): Linear(in_features=3072, out_features=768, bias=True)\n","          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","        )\n","      )\n","      (layernorm_embedding): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n","    )\n","  )\n","  (classification_head): MultimodalBartIntentClassification(\n","    (dense): Linear(in_features=768, out_features=768, bias=True)\n","    (dropout): Dropout(p=0.0, inplace=False)\n","    (out_proj): Linear(in_features=768, out_features=12, bias=True)\n","  )\n",")"]},"metadata":{},"execution_count":75}],"source":["model_intent_only"]},{"cell_type":"code","execution_count":76,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":31,"status":"ok","timestamp":1684514986775,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"n_N4HcMQ9hXX","outputId":"b208dc33-0bcb-41ec-ada5-4b7d95cc16fe"},"outputs":[{"output_type":"stream","name":"stdout","text":["Count :  0  name :  model.shared.weight\n","Count :  1  name :  model.encoder.embed_positions.weight\n","Count :  2  name :  model.encoder.layers.0.self_attn.k_proj.weight\n","Count :  3  name :  model.encoder.layers.0.self_attn.k_proj.bias\n","Count :  4  name :  model.encoder.layers.0.self_attn.v_proj.weight\n","Count :  5  name :  model.encoder.layers.0.self_attn.v_proj.bias\n","Count :  6  name :  model.encoder.layers.0.self_attn.q_proj.weight\n","Count :  7  name :  model.encoder.layers.0.self_attn.q_proj.bias\n","Count :  8  name :  model.encoder.layers.0.self_attn.out_proj.weight\n","Count :  9  name :  model.encoder.layers.0.self_attn.out_proj.bias\n","Count :  10  name :  model.encoder.layers.0.self_attn_layer_norm.weight\n","Count :  11  name :  model.encoder.layers.0.self_attn_layer_norm.bias\n","Count :  12  name :  model.encoder.layers.0.fc1.weight\n","Count :  13  name :  model.encoder.layers.0.fc1.bias\n","Count :  14  name :  model.encoder.layers.0.fc2.weight\n","Count :  15  name :  model.encoder.layers.0.fc2.bias\n","Count :  16  name :  model.encoder.layers.0.final_layer_norm.weight\n","Count :  17  name :  model.encoder.layers.0.final_layer_norm.bias\n","Count :  18  name :  model.encoder.layers.1.self_attn.k_proj.weight\n","Count :  19  name :  model.encoder.layers.1.self_attn.k_proj.bias\n","Count :  20  name :  model.encoder.layers.1.self_attn.v_proj.weight\n","Count :  21  name :  model.encoder.layers.1.self_attn.v_proj.bias\n","Count :  22  name :  model.encoder.layers.1.self_attn.q_proj.weight\n","Count :  23  name :  model.encoder.layers.1.self_attn.q_proj.bias\n","Count :  24  name :  model.encoder.layers.1.self_attn.out_proj.weight\n","Count :  25  name :  model.encoder.layers.1.self_attn.out_proj.bias\n","Count :  26  name :  model.encoder.layers.1.self_attn_layer_norm.weight\n","Count :  27  name :  model.encoder.layers.1.self_attn_layer_norm.bias\n","Count :  28  name :  model.encoder.layers.1.fc1.weight\n","Count :  29  name :  model.encoder.layers.1.fc1.bias\n","Count :  30  name :  model.encoder.layers.1.fc2.weight\n","Count :  31  name :  model.encoder.layers.1.fc2.bias\n","Count :  32  name :  model.encoder.layers.1.final_layer_norm.weight\n","Count :  33  name :  model.encoder.layers.1.final_layer_norm.bias\n","Count :  34  name :  model.encoder.layers.2.self_attn.k_proj.weight\n","Count :  35  name :  model.encoder.layers.2.self_attn.k_proj.bias\n","Count :  36  name :  model.encoder.layers.2.self_attn.v_proj.weight\n","Count :  37  name :  model.encoder.layers.2.self_attn.v_proj.bias\n","Count :  38  name :  model.encoder.layers.2.self_attn.q_proj.weight\n","Count :  39  name :  model.encoder.layers.2.self_attn.q_proj.bias\n","Count :  40  name :  model.encoder.layers.2.self_attn.out_proj.weight\n","Count :  41  name :  model.encoder.layers.2.self_attn.out_proj.bias\n","Count :  42  name :  model.encoder.layers.2.self_attn_layer_norm.weight\n","Count :  43  name :  model.encoder.layers.2.self_attn_layer_norm.bias\n","Count :  44  name :  model.encoder.layers.2.fc1.weight\n","Count :  45  name :  model.encoder.layers.2.fc1.bias\n","Count :  46  name :  model.encoder.layers.2.fc2.weight\n","Count :  47  name :  model.encoder.layers.2.fc2.bias\n","Count :  48  name :  model.encoder.layers.2.final_layer_norm.weight\n","Count :  49  name :  model.encoder.layers.2.final_layer_norm.bias\n","Count :  50  name :  model.encoder.layers.3.self_attn.k_proj.weight\n","Count :  51  name :  model.encoder.layers.3.self_attn.k_proj.bias\n","Count :  52  name :  model.encoder.layers.3.self_attn.v_proj.weight\n","Count :  53  name :  model.encoder.layers.3.self_attn.v_proj.bias\n","Count :  54  name :  model.encoder.layers.3.self_attn.q_proj.weight\n","Count :  55  name :  model.encoder.layers.3.self_attn.q_proj.bias\n","Count :  56  name :  model.encoder.layers.3.self_attn.out_proj.weight\n","Count :  57  name :  model.encoder.layers.3.self_attn.out_proj.bias\n","Count :  58  name :  model.encoder.layers.3.self_attn_layer_norm.weight\n","Count :  59  name :  model.encoder.layers.3.self_attn_layer_norm.bias\n","Count :  60  name :  model.encoder.layers.3.fc1.weight\n","Count :  61  name :  model.encoder.layers.3.fc1.bias\n","Count :  62  name :  model.encoder.layers.3.fc2.weight\n","Count :  63  name :  model.encoder.layers.3.fc2.bias\n","Count :  64  name :  model.encoder.layers.3.final_layer_norm.weight\n","Count :  65  name :  model.encoder.layers.3.final_layer_norm.bias\n","Count :  66  name :  model.encoder.layers.4.self_attn.k_proj.weight\n","Count :  67  name :  model.encoder.layers.4.self_attn.k_proj.bias\n","Count :  68  name :  model.encoder.layers.4.self_attn.v_proj.weight\n","Count :  69  name :  model.encoder.layers.4.self_attn.v_proj.bias\n","Count :  70  name :  model.encoder.layers.4.self_attn.q_proj.weight\n","Count :  71  name :  model.encoder.layers.4.self_attn.q_proj.bias\n","Count :  72  name :  model.encoder.layers.4.self_attn.out_proj.weight\n","Count :  73  name :  model.encoder.layers.4.self_attn.out_proj.bias\n","Count :  74  name :  model.encoder.layers.4.self_attn_layer_norm.weight\n","Count :  75  name :  model.encoder.layers.4.self_attn_layer_norm.bias\n","Count :  76  name :  model.encoder.layers.4.fc1.weight\n","Count :  77  name :  model.encoder.layers.4.fc1.bias\n","Count :  78  name :  model.encoder.layers.4.fc2.weight\n","Count :  79  name :  model.encoder.layers.4.fc2.bias\n","Count :  80  name :  model.encoder.layers.4.final_layer_norm.weight\n","Count :  81  name :  model.encoder.layers.4.final_layer_norm.bias\n","Count :  82  name :  model.encoder.layers.5.self_attn.k_proj.weight\n","Count :  83  name :  model.encoder.layers.5.self_attn.k_proj.bias\n","Count :  84  name :  model.encoder.layers.5.self_attn.v_proj.weight\n","Count :  85  name :  model.encoder.layers.5.self_attn.v_proj.bias\n","Count :  86  name :  model.encoder.layers.5.self_attn.q_proj.weight\n","Count :  87  name :  model.encoder.layers.5.self_attn.q_proj.bias\n","Count :  88  name :  model.encoder.layers.5.self_attn.out_proj.weight\n","Count :  89  name :  model.encoder.layers.5.self_attn.out_proj.bias\n","Count :  90  name :  model.encoder.layers.5.self_attn_layer_norm.weight\n","Count :  91  name :  model.encoder.layers.5.self_attn_layer_norm.bias\n","Count :  92  name :  model.encoder.layers.5.fc1.weight\n","Count :  93  name :  model.encoder.layers.5.fc1.bias\n","Count :  94  name :  model.encoder.layers.5.fc2.weight\n","Count :  95  name :  model.encoder.layers.5.fc2.bias\n","Count :  96  name :  model.encoder.layers.5.final_layer_norm.weight\n","Count :  97  name :  model.encoder.layers.5.final_layer_norm.bias\n","Count :  98  name :  model.encoder.layernorm_embedding.weight\n","Count :  99  name :  model.encoder.layernorm_embedding.bias\n","Count :  100  name :  model.encoder.MAF_layer4.acoustic_context_transform.weight\n","Count :  101  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_weight\n","Count :  102  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_bias\n","Count :  103  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.weight\n","Count :  104  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.bias\n","Count :  105  name :  model.encoder.MAF_layer4.acoustic_context_attention.u_k.weight\n","Count :  106  name :  model.encoder.MAF_layer4.acoustic_context_attention.w1_k.weight\n","Count :  107  name :  model.encoder.MAF_layer4.acoustic_context_attention.w2_k.weight\n","Count :  108  name :  model.encoder.MAF_layer4.acoustic_context_attention.u_v.weight\n","Count :  109  name :  model.encoder.MAF_layer4.acoustic_context_attention.w1_v.weight\n","Count :  110  name :  model.encoder.MAF_layer4.acoustic_context_attention.w2_v.weight\n","Count :  111  name :  model.encoder.MAF_layer4.acoustic_gate.weight\n","Count :  112  name :  model.encoder.MAF_layer4.acoustic_gate.bias\n","Count :  113  name :  model.encoder.MAF_layer4.final_layer_norm.weight\n","Count :  114  name :  model.encoder.MAF_layer4.final_layer_norm.bias\n","Count :  115  name :  model.encoder.MAF_layer5.visual_context_transform.weight\n","Count :  116  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_weight\n","Count :  117  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_bias\n","Count :  118  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.weight\n","Count :  119  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.bias\n","Count :  120  name :  model.encoder.MAF_layer5.visual_context_attention.u_k.weight\n","Count :  121  name :  model.encoder.MAF_layer5.visual_context_attention.w1_k.weight\n","Count :  122  name :  model.encoder.MAF_layer5.visual_context_attention.w2_k.weight\n","Count :  123  name :  model.encoder.MAF_layer5.visual_context_attention.u_v.weight\n","Count :  124  name :  model.encoder.MAF_layer5.visual_context_attention.w1_v.weight\n","Count :  125  name :  model.encoder.MAF_layer5.visual_context_attention.w2_v.weight\n","Count :  126  name :  model.encoder.MAF_layer5.visual_gate.weight\n","Count :  127  name :  model.encoder.MAF_layer5.visual_gate.bias\n","Count :  128  name :  model.encoder.MAF_layer5.final_layer_norm.weight\n","Count :  129  name :  model.encoder.MAF_layer5.final_layer_norm.bias\n","Count :  130  name :  model.decoder.embed_positions.weight\n","Count :  131  name :  model.decoder.layers.0.self_attn.k_proj.weight\n","Count :  132  name :  model.decoder.layers.0.self_attn.k_proj.bias\n","Count :  133  name :  model.decoder.layers.0.self_attn.v_proj.weight\n","Count :  134  name :  model.decoder.layers.0.self_attn.v_proj.bias\n","Count :  135  name :  model.decoder.layers.0.self_attn.q_proj.weight\n","Count :  136  name :  model.decoder.layers.0.self_attn.q_proj.bias\n","Count :  137  name :  model.decoder.layers.0.self_attn.out_proj.weight\n","Count :  138  name :  model.decoder.layers.0.self_attn.out_proj.bias\n","Count :  139  name :  model.decoder.layers.0.self_attn_layer_norm.weight\n","Count :  140  name :  model.decoder.layers.0.self_attn_layer_norm.bias\n","Count :  141  name :  model.decoder.layers.0.encoder_attn.k_proj.weight\n","Count :  142  name :  model.decoder.layers.0.encoder_attn.k_proj.bias\n","Count :  143  name :  model.decoder.layers.0.encoder_attn.v_proj.weight\n","Count :  144  name :  model.decoder.layers.0.encoder_attn.v_proj.bias\n","Count :  145  name :  model.decoder.layers.0.encoder_attn.q_proj.weight\n","Count :  146  name :  model.decoder.layers.0.encoder_attn.q_proj.bias\n","Count :  147  name :  model.decoder.layers.0.encoder_attn.out_proj.weight\n","Count :  148  name :  model.decoder.layers.0.encoder_attn.out_proj.bias\n","Count :  149  name :  model.decoder.layers.0.encoder_attn_layer_norm.weight\n","Count :  150  name :  model.decoder.layers.0.encoder_attn_layer_norm.bias\n","Count :  151  name :  model.decoder.layers.0.fc1.weight\n","Count :  152  name :  model.decoder.layers.0.fc1.bias\n","Count :  153  name :  model.decoder.layers.0.fc2.weight\n","Count :  154  name :  model.decoder.layers.0.fc2.bias\n","Count :  155  name :  model.decoder.layers.0.final_layer_norm.weight\n","Count :  156  name :  model.decoder.layers.0.final_layer_norm.bias\n","Count :  157  name :  model.decoder.layers.1.self_attn.k_proj.weight\n","Count :  158  name :  model.decoder.layers.1.self_attn.k_proj.bias\n","Count :  159  name :  model.decoder.layers.1.self_attn.v_proj.weight\n","Count :  160  name :  model.decoder.layers.1.self_attn.v_proj.bias\n","Count :  161  name :  model.decoder.layers.1.self_attn.q_proj.weight\n","Count :  162  name :  model.decoder.layers.1.self_attn.q_proj.bias\n","Count :  163  name :  model.decoder.layers.1.self_attn.out_proj.weight\n","Count :  164  name :  model.decoder.layers.1.self_attn.out_proj.bias\n","Count :  165  name :  model.decoder.layers.1.self_attn_layer_norm.weight\n","Count :  166  name :  model.decoder.layers.1.self_attn_layer_norm.bias\n","Count :  167  name :  model.decoder.layers.1.encoder_attn.k_proj.weight\n","Count :  168  name :  model.decoder.layers.1.encoder_attn.k_proj.bias\n","Count :  169  name :  model.decoder.layers.1.encoder_attn.v_proj.weight\n","Count :  170  name :  model.decoder.layers.1.encoder_attn.v_proj.bias\n","Count :  171  name :  model.decoder.layers.1.encoder_attn.q_proj.weight\n","Count :  172  name :  model.decoder.layers.1.encoder_attn.q_proj.bias\n","Count :  173  name :  model.decoder.layers.1.encoder_attn.out_proj.weight\n","Count :  174  name :  model.decoder.layers.1.encoder_attn.out_proj.bias\n","Count :  175  name :  model.decoder.layers.1.encoder_attn_layer_norm.weight\n","Count :  176  name :  model.decoder.layers.1.encoder_attn_layer_norm.bias\n","Count :  177  name :  model.decoder.layers.1.fc1.weight\n","Count :  178  name :  model.decoder.layers.1.fc1.bias\n","Count :  179  name :  model.decoder.layers.1.fc2.weight\n","Count :  180  name :  model.decoder.layers.1.fc2.bias\n","Count :  181  name :  model.decoder.layers.1.final_layer_norm.weight\n","Count :  182  name :  model.decoder.layers.1.final_layer_norm.bias\n","Count :  183  name :  model.decoder.layers.2.self_attn.k_proj.weight\n","Count :  184  name :  model.decoder.layers.2.self_attn.k_proj.bias\n","Count :  185  name :  model.decoder.layers.2.self_attn.v_proj.weight\n","Count :  186  name :  model.decoder.layers.2.self_attn.v_proj.bias\n","Count :  187  name :  model.decoder.layers.2.self_attn.q_proj.weight\n","Count :  188  name :  model.decoder.layers.2.self_attn.q_proj.bias\n","Count :  189  name :  model.decoder.layers.2.self_attn.out_proj.weight\n","Count :  190  name :  model.decoder.layers.2.self_attn.out_proj.bias\n","Count :  191  name :  model.decoder.layers.2.self_attn_layer_norm.weight\n","Count :  192  name :  model.decoder.layers.2.self_attn_layer_norm.bias\n","Count :  193  name :  model.decoder.layers.2.encoder_attn.k_proj.weight\n","Count :  194  name :  model.decoder.layers.2.encoder_attn.k_proj.bias\n","Count :  195  name :  model.decoder.layers.2.encoder_attn.v_proj.weight\n","Count :  196  name :  model.decoder.layers.2.encoder_attn.v_proj.bias\n","Count :  197  name :  model.decoder.layers.2.encoder_attn.q_proj.weight\n","Count :  198  name :  model.decoder.layers.2.encoder_attn.q_proj.bias\n","Count :  199  name :  model.decoder.layers.2.encoder_attn.out_proj.weight\n","Count :  200  name :  model.decoder.layers.2.encoder_attn.out_proj.bias\n","Count :  201  name :  model.decoder.layers.2.encoder_attn_layer_norm.weight\n","Count :  202  name :  model.decoder.layers.2.encoder_attn_layer_norm.bias\n","Count :  203  name :  model.decoder.layers.2.fc1.weight\n","Count :  204  name :  model.decoder.layers.2.fc1.bias\n","Count :  205  name :  model.decoder.layers.2.fc2.weight\n","Count :  206  name :  model.decoder.layers.2.fc2.bias\n","Count :  207  name :  model.decoder.layers.2.final_layer_norm.weight\n","Count :  208  name :  model.decoder.layers.2.final_layer_norm.bias\n","Count :  209  name :  model.decoder.layers.3.self_attn.k_proj.weight\n","Count :  210  name :  model.decoder.layers.3.self_attn.k_proj.bias\n","Count :  211  name :  model.decoder.layers.3.self_attn.v_proj.weight\n","Count :  212  name :  model.decoder.layers.3.self_attn.v_proj.bias\n","Count :  213  name :  model.decoder.layers.3.self_attn.q_proj.weight\n","Count :  214  name :  model.decoder.layers.3.self_attn.q_proj.bias\n","Count :  215  name :  model.decoder.layers.3.self_attn.out_proj.weight\n","Count :  216  name :  model.decoder.layers.3.self_attn.out_proj.bias\n","Count :  217  name :  model.decoder.layers.3.self_attn_layer_norm.weight\n","Count :  218  name :  model.decoder.layers.3.self_attn_layer_norm.bias\n","Count :  219  name :  model.decoder.layers.3.encoder_attn.k_proj.weight\n","Count :  220  name :  model.decoder.layers.3.encoder_attn.k_proj.bias\n","Count :  221  name :  model.decoder.layers.3.encoder_attn.v_proj.weight\n","Count :  222  name :  model.decoder.layers.3.encoder_attn.v_proj.bias\n","Count :  223  name :  model.decoder.layers.3.encoder_attn.q_proj.weight\n","Count :  224  name :  model.decoder.layers.3.encoder_attn.q_proj.bias\n","Count :  225  name :  model.decoder.layers.3.encoder_attn.out_proj.weight\n","Count :  226  name :  model.decoder.layers.3.encoder_attn.out_proj.bias\n","Count :  227  name :  model.decoder.layers.3.encoder_attn_layer_norm.weight\n","Count :  228  name :  model.decoder.layers.3.encoder_attn_layer_norm.bias\n","Count :  229  name :  model.decoder.layers.3.fc1.weight\n","Count :  230  name :  model.decoder.layers.3.fc1.bias\n","Count :  231  name :  model.decoder.layers.3.fc2.weight\n","Count :  232  name :  model.decoder.layers.3.fc2.bias\n","Count :  233  name :  model.decoder.layers.3.final_layer_norm.weight\n","Count :  234  name :  model.decoder.layers.3.final_layer_norm.bias\n","Count :  235  name :  model.decoder.layers.4.self_attn.k_proj.weight\n","Count :  236  name :  model.decoder.layers.4.self_attn.k_proj.bias\n","Count :  237  name :  model.decoder.layers.4.self_attn.v_proj.weight\n","Count :  238  name :  model.decoder.layers.4.self_attn.v_proj.bias\n","Count :  239  name :  model.decoder.layers.4.self_attn.q_proj.weight\n","Count :  240  name :  model.decoder.layers.4.self_attn.q_proj.bias\n","Count :  241  name :  model.decoder.layers.4.self_attn.out_proj.weight\n","Count :  242  name :  model.decoder.layers.4.self_attn.out_proj.bias\n","Count :  243  name :  model.decoder.layers.4.self_attn_layer_norm.weight\n","Count :  244  name :  model.decoder.layers.4.self_attn_layer_norm.bias\n","Count :  245  name :  model.decoder.layers.4.encoder_attn.k_proj.weight\n","Count :  246  name :  model.decoder.layers.4.encoder_attn.k_proj.bias\n","Count :  247  name :  model.decoder.layers.4.encoder_attn.v_proj.weight\n","Count :  248  name :  model.decoder.layers.4.encoder_attn.v_proj.bias\n","Count :  249  name :  model.decoder.layers.4.encoder_attn.q_proj.weight\n","Count :  250  name :  model.decoder.layers.4.encoder_attn.q_proj.bias\n","Count :  251  name :  model.decoder.layers.4.encoder_attn.out_proj.weight\n","Count :  252  name :  model.decoder.layers.4.encoder_attn.out_proj.bias\n","Count :  253  name :  model.decoder.layers.4.encoder_attn_layer_norm.weight\n","Count :  254  name :  model.decoder.layers.4.encoder_attn_layer_norm.bias\n","Count :  255  name :  model.decoder.layers.4.fc1.weight\n","Count :  256  name :  model.decoder.layers.4.fc1.bias\n","Count :  257  name :  model.decoder.layers.4.fc2.weight\n","Count :  258  name :  model.decoder.layers.4.fc2.bias\n","Count :  259  name :  model.decoder.layers.4.final_layer_norm.weight\n","Count :  260  name :  model.decoder.layers.4.final_layer_norm.bias\n","Count :  261  name :  model.decoder.layers.5.self_attn.k_proj.weight\n","Count :  262  name :  model.decoder.layers.5.self_attn.k_proj.bias\n","Count :  263  name :  model.decoder.layers.5.self_attn.v_proj.weight\n","Count :  264  name :  model.decoder.layers.5.self_attn.v_proj.bias\n","Count :  265  name :  model.decoder.layers.5.self_attn.q_proj.weight\n","Count :  266  name :  model.decoder.layers.5.self_attn.q_proj.bias\n","Count :  267  name :  model.decoder.layers.5.self_attn.out_proj.weight\n","Count :  268  name :  model.decoder.layers.5.self_attn.out_proj.bias\n","Count :  269  name :  model.decoder.layers.5.self_attn_layer_norm.weight\n","Count :  270  name :  model.decoder.layers.5.self_attn_layer_norm.bias\n","Count :  271  name :  model.decoder.layers.5.encoder_attn.k_proj.weight\n","Count :  272  name :  model.decoder.layers.5.encoder_attn.k_proj.bias\n","Count :  273  name :  model.decoder.layers.5.encoder_attn.v_proj.weight\n","Count :  274  name :  model.decoder.layers.5.encoder_attn.v_proj.bias\n","Count :  275  name :  model.decoder.layers.5.encoder_attn.q_proj.weight\n","Count :  276  name :  model.decoder.layers.5.encoder_attn.q_proj.bias\n","Count :  277  name :  model.decoder.layers.5.encoder_attn.out_proj.weight\n","Count :  278  name :  model.decoder.layers.5.encoder_attn.out_proj.bias\n","Count :  279  name :  model.decoder.layers.5.encoder_attn_layer_norm.weight\n","Count :  280  name :  model.decoder.layers.5.encoder_attn_layer_norm.bias\n","Count :  281  name :  model.decoder.layers.5.fc1.weight\n","Count :  282  name :  model.decoder.layers.5.fc1.bias\n","Count :  283  name :  model.decoder.layers.5.fc2.weight\n","Count :  284  name :  model.decoder.layers.5.fc2.bias\n","Count :  285  name :  model.decoder.layers.5.final_layer_norm.weight\n","Count :  286  name :  model.decoder.layers.5.final_layer_norm.bias\n","Count :  287  name :  model.decoder.layernorm_embedding.weight\n","Count :  288  name :  model.decoder.layernorm_embedding.bias\n","Count :  289  name :  classification_head.dense.weight\n","Count :  290  name :  classification_head.dense.bias\n","Count :  291  name :  classification_head.out_proj.weight\n","Count :  292  name :  classification_head.out_proj.bias\n"]}],"source":["cnt = 0\n","for name, param in model_sarcasm_only.named_parameters():\n","    print(\"Count : \", cnt, \" name : \", name)\n","    cnt+=1"]},{"cell_type":"code","execution_count":77,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":25,"status":"ok","timestamp":1684514986775,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"feA-7jSwIBQO","outputId":"d25e97ac-ca5e-42eb-a044-ef8e4f6a197b"},"outputs":[{"output_type":"stream","name":"stdout","text":["Total trainanable parameters :  0.0\n"]}],"source":["cnt = 0\n","for name, param in model_sarcasm_only.named_parameters():\n","\n","    param.requires_grad = False    \n","    cnt+=1\n","        \n","\n","num_param = sum(p.numel() for p in model_sarcasm_only.parameters() if p.requires_grad)\n","print(\"Total trainanable parameters : \", num_param/1e6)"]},{"cell_type":"code","execution_count":78,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":792,"status":"ok","timestamp":1684514987550,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"ywwajnJZ8ql3","outputId":"2a2aee60-8761-4bf7-edb8-02c58a0f8612"},"outputs":[{"output_type":"stream","name":"stdout","text":["Count :  0  name :  model.shared.weight\n","Count :  1  name :  model.encoder.embed_positions.weight\n","Count :  2  name :  model.encoder.layers.0.self_attn.k_proj.weight\n","Count :  3  name :  model.encoder.layers.0.self_attn.k_proj.bias\n","Count :  4  name :  model.encoder.layers.0.self_attn.v_proj.weight\n","Count :  5  name :  model.encoder.layers.0.self_attn.v_proj.bias\n","Count :  6  name :  model.encoder.layers.0.self_attn.q_proj.weight\n","Count :  7  name :  model.encoder.layers.0.self_attn.q_proj.bias\n","Count :  8  name :  model.encoder.layers.0.self_attn.out_proj.weight\n","Count :  9  name :  model.encoder.layers.0.self_attn.out_proj.bias\n","Count :  10  name :  model.encoder.layers.0.self_attn_layer_norm.weight\n","Count :  11  name :  model.encoder.layers.0.self_attn_layer_norm.bias\n","Count :  12  name :  model.encoder.layers.0.fc1.weight\n","Count :  13  name :  model.encoder.layers.0.fc1.bias\n","Count :  14  name :  model.encoder.layers.0.fc2.weight\n","Count :  15  name :  model.encoder.layers.0.fc2.bias\n","Count :  16  name :  model.encoder.layers.0.final_layer_norm.weight\n","Count :  17  name :  model.encoder.layers.0.final_layer_norm.bias\n","Count :  18  name :  model.encoder.layers.1.self_attn.k_proj.weight\n","Count :  19  name :  model.encoder.layers.1.self_attn.k_proj.bias\n","Count :  20  name :  model.encoder.layers.1.self_attn.v_proj.weight\n","Count :  21  name :  model.encoder.layers.1.self_attn.v_proj.bias\n","Count :  22  name :  model.encoder.layers.1.self_attn.q_proj.weight\n","Count :  23  name :  model.encoder.layers.1.self_attn.q_proj.bias\n","Count :  24  name :  model.encoder.layers.1.self_attn.out_proj.weight\n","Count :  25  name :  model.encoder.layers.1.self_attn.out_proj.bias\n","Count :  26  name :  model.encoder.layers.1.self_attn_layer_norm.weight\n","Count :  27  name :  model.encoder.layers.1.self_attn_layer_norm.bias\n","Count :  28  name :  model.encoder.layers.1.fc1.weight\n","Count :  29  name :  model.encoder.layers.1.fc1.bias\n","Count :  30  name :  model.encoder.layers.1.fc2.weight\n","Count :  31  name :  model.encoder.layers.1.fc2.bias\n","Count :  32  name :  model.encoder.layers.1.final_layer_norm.weight\n","Count :  33  name :  model.encoder.layers.1.final_layer_norm.bias\n","Count :  34  name :  model.encoder.layers.2.self_attn.k_proj.weight\n","Count :  35  name :  model.encoder.layers.2.self_attn.k_proj.bias\n","Count :  36  name :  model.encoder.layers.2.self_attn.v_proj.weight\n","Count :  37  name :  model.encoder.layers.2.self_attn.v_proj.bias\n","Count :  38  name :  model.encoder.layers.2.self_attn.q_proj.weight\n","Count :  39  name :  model.encoder.layers.2.self_attn.q_proj.bias\n","Count :  40  name :  model.encoder.layers.2.self_attn.out_proj.weight\n","Count :  41  name :  model.encoder.layers.2.self_attn.out_proj.bias\n","Count :  42  name :  model.encoder.layers.2.self_attn_layer_norm.weight\n","Count :  43  name :  model.encoder.layers.2.self_attn_layer_norm.bias\n","Count :  44  name :  model.encoder.layers.2.fc1.weight\n","Count :  45  name :  model.encoder.layers.2.fc1.bias\n","Count :  46  name :  model.encoder.layers.2.fc2.weight\n","Count :  47  name :  model.encoder.layers.2.fc2.bias\n","Count :  48  name :  model.encoder.layers.2.final_layer_norm.weight\n","Count :  49  name :  model.encoder.layers.2.final_layer_norm.bias\n","Count :  50  name :  model.encoder.layers.3.self_attn.k_proj.weight\n","Count :  51  name :  model.encoder.layers.3.self_attn.k_proj.bias\n","Count :  52  name :  model.encoder.layers.3.self_attn.v_proj.weight\n","Count :  53  name :  model.encoder.layers.3.self_attn.v_proj.bias\n","Count :  54  name :  model.encoder.layers.3.self_attn.q_proj.weight\n","Count :  55  name :  model.encoder.layers.3.self_attn.q_proj.bias\n","Count :  56  name :  model.encoder.layers.3.self_attn.out_proj.weight\n","Count :  57  name :  model.encoder.layers.3.self_attn.out_proj.bias\n","Count :  58  name :  model.encoder.layers.3.self_attn_layer_norm.weight\n","Count :  59  name :  model.encoder.layers.3.self_attn_layer_norm.bias\n","Count :  60  name :  model.encoder.layers.3.fc1.weight\n","Count :  61  name :  model.encoder.layers.3.fc1.bias\n","Count :  62  name :  model.encoder.layers.3.fc2.weight\n","Count :  63  name :  model.encoder.layers.3.fc2.bias\n","Count :  64  name :  model.encoder.layers.3.final_layer_norm.weight\n","Count :  65  name :  model.encoder.layers.3.final_layer_norm.bias\n","Count :  66  name :  model.encoder.layers.4.self_attn.k_proj.weight\n","Count :  67  name :  model.encoder.layers.4.self_attn.k_proj.bias\n","Count :  68  name :  model.encoder.layers.4.self_attn.v_proj.weight\n","Count :  69  name :  model.encoder.layers.4.self_attn.v_proj.bias\n","Count :  70  name :  model.encoder.layers.4.self_attn.q_proj.weight\n","Count :  71  name :  model.encoder.layers.4.self_attn.q_proj.bias\n","Count :  72  name :  model.encoder.layers.4.self_attn.out_proj.weight\n","Count :  73  name :  model.encoder.layers.4.self_attn.out_proj.bias\n","Count :  74  name :  model.encoder.layers.4.self_attn_layer_norm.weight\n","Count :  75  name :  model.encoder.layers.4.self_attn_layer_norm.bias\n","Count :  76  name :  model.encoder.layers.4.fc1.weight\n","Count :  77  name :  model.encoder.layers.4.fc1.bias\n","Count :  78  name :  model.encoder.layers.4.fc2.weight\n","Count :  79  name :  model.encoder.layers.4.fc2.bias\n","Count :  80  name :  model.encoder.layers.4.final_layer_norm.weight\n","Count :  81  name :  model.encoder.layers.4.final_layer_norm.bias\n","Count :  82  name :  model.encoder.layers.5.self_attn.k_proj.weight\n","Count :  83  name :  model.encoder.layers.5.self_attn.k_proj.bias\n","Count :  84  name :  model.encoder.layers.5.self_attn.v_proj.weight\n","Count :  85  name :  model.encoder.layers.5.self_attn.v_proj.bias\n","Count :  86  name :  model.encoder.layers.5.self_attn.q_proj.weight\n","Count :  87  name :  model.encoder.layers.5.self_attn.q_proj.bias\n","Count :  88  name :  model.encoder.layers.5.self_attn.out_proj.weight\n","Count :  89  name :  model.encoder.layers.5.self_attn.out_proj.bias\n","Count :  90  name :  model.encoder.layers.5.self_attn_layer_norm.weight\n","Count :  91  name :  model.encoder.layers.5.self_attn_layer_norm.bias\n","Count :  92  name :  model.encoder.layers.5.fc1.weight\n","Count :  93  name :  model.encoder.layers.5.fc1.bias\n","Count :  94  name :  model.encoder.layers.5.fc2.weight\n","Count :  95  name :  model.encoder.layers.5.fc2.bias\n","Count :  96  name :  model.encoder.layers.5.final_layer_norm.weight\n","Count :  97  name :  model.encoder.layers.5.final_layer_norm.bias\n","Count :  98  name :  model.encoder.layernorm_embedding.weight\n","Count :  99  name :  model.encoder.layernorm_embedding.bias\n","Count :  100  name :  model.encoder.MAF_layer4.acoustic_context_transform.weight\n","Count :  101  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_weight\n","Count :  102  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.in_proj_bias\n","Count :  103  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.weight\n","Count :  104  name :  model.encoder.MAF_layer4.acoustic_context_attention.attention_layer.out_proj.bias\n","Count :  105  name :  model.encoder.MAF_layer4.acoustic_context_attention.u_k.weight\n","Count :  106  name :  model.encoder.MAF_layer4.acoustic_context_attention.w1_k.weight\n","Count :  107  name :  model.encoder.MAF_layer4.acoustic_context_attention.w2_k.weight\n","Count :  108  name :  model.encoder.MAF_layer4.acoustic_context_attention.u_v.weight\n","Count :  109  name :  model.encoder.MAF_layer4.acoustic_context_attention.w1_v.weight\n","Count :  110  name :  model.encoder.MAF_layer4.acoustic_context_attention.w2_v.weight\n","Count :  111  name :  model.encoder.MAF_layer4.acoustic_gate.weight\n","Count :  112  name :  model.encoder.MAF_layer4.acoustic_gate.bias\n","Count :  113  name :  model.encoder.MAF_layer4.final_layer_norm.weight\n","Count :  114  name :  model.encoder.MAF_layer4.final_layer_norm.bias\n","Count :  115  name :  model.encoder.MAF_layer5.visual_context_transform.weight\n","Count :  116  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_weight\n","Count :  117  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.in_proj_bias\n","Count :  118  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.weight\n","Count :  119  name :  model.encoder.MAF_layer5.visual_context_attention.attention_layer.out_proj.bias\n","Count :  120  name :  model.encoder.MAF_layer5.visual_context_attention.u_k.weight\n","Count :  121  name :  model.encoder.MAF_layer5.visual_context_attention.w1_k.weight\n","Count :  122  name :  model.encoder.MAF_layer5.visual_context_attention.w2_k.weight\n","Count :  123  name :  model.encoder.MAF_layer5.visual_context_attention.u_v.weight\n","Count :  124  name :  model.encoder.MAF_layer5.visual_context_attention.w1_v.weight\n","Count :  125  name :  model.encoder.MAF_layer5.visual_context_attention.w2_v.weight\n","Count :  126  name :  model.encoder.MAF_layer5.visual_gate.weight\n","Count :  127  name :  model.encoder.MAF_layer5.visual_gate.bias\n","Count :  128  name :  model.encoder.MAF_layer5.final_layer_norm.weight\n","Count :  129  name :  model.encoder.MAF_layer5.final_layer_norm.bias\n","Count :  130  name :  model.decoder.embed_positions.weight\n","Count :  131  name :  model.decoder.layers.0.self_attn.k_proj.weight\n","Count :  132  name :  model.decoder.layers.0.self_attn.k_proj.bias\n","Count :  133  name :  model.decoder.layers.0.self_attn.v_proj.weight\n","Count :  134  name :  model.decoder.layers.0.self_attn.v_proj.bias\n","Count :  135  name :  model.decoder.layers.0.self_attn.q_proj.weight\n","Count :  136  name :  model.decoder.layers.0.self_attn.q_proj.bias\n","Count :  137  name :  model.decoder.layers.0.self_attn.out_proj.weight\n","Count :  138  name :  model.decoder.layers.0.self_attn.out_proj.bias\n","Count :  139  name :  model.decoder.layers.0.self_attn_layer_norm.weight\n","Count :  140  name :  model.decoder.layers.0.self_attn_layer_norm.bias\n","Count :  141  name :  model.decoder.layers.0.encoder_attn.k_proj.weight\n","Count :  142  name :  model.decoder.layers.0.encoder_attn.k_proj.bias\n","Count :  143  name :  model.decoder.layers.0.encoder_attn.v_proj.weight\n","Count :  144  name :  model.decoder.layers.0.encoder_attn.v_proj.bias\n","Count :  145  name :  model.decoder.layers.0.encoder_attn.q_proj.weight\n","Count :  146  name :  model.decoder.layers.0.encoder_attn.q_proj.bias\n","Count :  147  name :  model.decoder.layers.0.encoder_attn.out_proj.weight\n","Count :  148  name :  model.decoder.layers.0.encoder_attn.out_proj.bias\n","Count :  149  name :  model.decoder.layers.0.encoder_attn_layer_norm.weight\n","Count :  150  name :  model.decoder.layers.0.encoder_attn_layer_norm.bias\n","Count :  151  name :  model.decoder.layers.0.fc1.weight\n","Count :  152  name :  model.decoder.layers.0.fc1.bias\n","Count :  153  name :  model.decoder.layers.0.fc2.weight\n","Count :  154  name :  model.decoder.layers.0.fc2.bias\n","Count :  155  name :  model.decoder.layers.0.final_layer_norm.weight\n","Count :  156  name :  model.decoder.layers.0.final_layer_norm.bias\n","Count :  157  name :  model.decoder.layers.1.self_attn.k_proj.weight\n","Count :  158  name :  model.decoder.layers.1.self_attn.k_proj.bias\n","Count :  159  name :  model.decoder.layers.1.self_attn.v_proj.weight\n","Count :  160  name :  model.decoder.layers.1.self_attn.v_proj.bias\n","Count :  161  name :  model.decoder.layers.1.self_attn.q_proj.weight\n","Count :  162  name :  model.decoder.layers.1.self_attn.q_proj.bias\n","Count :  163  name :  model.decoder.layers.1.self_attn.out_proj.weight\n","Count :  164  name :  model.decoder.layers.1.self_attn.out_proj.bias\n","Count :  165  name :  model.decoder.layers.1.self_attn_layer_norm.weight\n","Count :  166  name :  model.decoder.layers.1.self_attn_layer_norm.bias\n","Count :  167  name :  model.decoder.layers.1.encoder_attn.k_proj.weight\n","Count :  168  name :  model.decoder.layers.1.encoder_attn.k_proj.bias\n","Count :  169  name :  model.decoder.layers.1.encoder_attn.v_proj.weight\n","Count :  170  name :  model.decoder.layers.1.encoder_attn.v_proj.bias\n","Count :  171  name :  model.decoder.layers.1.encoder_attn.q_proj.weight\n","Count :  172  name :  model.decoder.layers.1.encoder_attn.q_proj.bias\n","Count :  173  name :  model.decoder.layers.1.encoder_attn.out_proj.weight\n","Count :  174  name :  model.decoder.layers.1.encoder_attn.out_proj.bias\n","Count :  175  name :  model.decoder.layers.1.encoder_attn_layer_norm.weight\n","Count :  176  name :  model.decoder.layers.1.encoder_attn_layer_norm.bias\n","Count :  177  name :  model.decoder.layers.1.fc1.weight\n","Count :  178  name :  model.decoder.layers.1.fc1.bias\n","Count :  179  name :  model.decoder.layers.1.fc2.weight\n","Count :  180  name :  model.decoder.layers.1.fc2.bias\n","Count :  181  name :  model.decoder.layers.1.final_layer_norm.weight\n","Count :  182  name :  model.decoder.layers.1.final_layer_norm.bias\n","Count :  183  name :  model.decoder.layers.2.self_attn.k_proj.weight\n","Count :  184  name :  model.decoder.layers.2.self_attn.k_proj.bias\n","Count :  185  name :  model.decoder.layers.2.self_attn.v_proj.weight\n","Count :  186  name :  model.decoder.layers.2.self_attn.v_proj.bias\n","Count :  187  name :  model.decoder.layers.2.self_attn.q_proj.weight\n","Count :  188  name :  model.decoder.layers.2.self_attn.q_proj.bias\n","Count :  189  name :  model.decoder.layers.2.self_attn.out_proj.weight\n","Count :  190  name :  model.decoder.layers.2.self_attn.out_proj.bias\n","Count :  191  name :  model.decoder.layers.2.self_attn_layer_norm.weight\n","Count :  192  name :  model.decoder.layers.2.self_attn_layer_norm.bias\n","Count :  193  name :  model.decoder.layers.2.encoder_attn.k_proj.weight\n","Count :  194  name :  model.decoder.layers.2.encoder_attn.k_proj.bias\n","Count :  195  name :  model.decoder.layers.2.encoder_attn.v_proj.weight\n","Count :  196  name :  model.decoder.layers.2.encoder_attn.v_proj.bias\n","Count :  197  name :  model.decoder.layers.2.encoder_attn.q_proj.weight\n","Count :  198  name :  model.decoder.layers.2.encoder_attn.q_proj.bias\n","Count :  199  name :  model.decoder.layers.2.encoder_attn.out_proj.weight\n","Count :  200  name :  model.decoder.layers.2.encoder_attn.out_proj.bias\n","Count :  201  name :  model.decoder.layers.2.encoder_attn_layer_norm.weight\n","Count :  202  name :  model.decoder.layers.2.encoder_attn_layer_norm.bias\n","Count :  203  name :  model.decoder.layers.2.fc1.weight\n","Count :  204  name :  model.decoder.layers.2.fc1.bias\n","Count :  205  name :  model.decoder.layers.2.fc2.weight\n","Count :  206  name :  model.decoder.layers.2.fc2.bias\n","Count :  207  name :  model.decoder.layers.2.final_layer_norm.weight\n","Count :  208  name :  model.decoder.layers.2.final_layer_norm.bias\n","Count :  209  name :  model.decoder.layers.3.self_attn.k_proj.weight\n","Count :  210  name :  model.decoder.layers.3.self_attn.k_proj.bias\n","Count :  211  name :  model.decoder.layers.3.self_attn.v_proj.weight\n","Count :  212  name :  model.decoder.layers.3.self_attn.v_proj.bias\n","Count :  213  name :  model.decoder.layers.3.self_attn.q_proj.weight\n","Count :  214  name :  model.decoder.layers.3.self_attn.q_proj.bias\n","Count :  215  name :  model.decoder.layers.3.self_attn.out_proj.weight\n","Count :  216  name :  model.decoder.layers.3.self_attn.out_proj.bias\n","Count :  217  name :  model.decoder.layers.3.self_attn_layer_norm.weight\n","Count :  218  name :  model.decoder.layers.3.self_attn_layer_norm.bias\n","Count :  219  name :  model.decoder.layers.3.encoder_attn.k_proj.weight\n","Count :  220  name :  model.decoder.layers.3.encoder_attn.k_proj.bias\n","Count :  221  name :  model.decoder.layers.3.encoder_attn.v_proj.weight\n","Count :  222  name :  model.decoder.layers.3.encoder_attn.v_proj.bias\n","Count :  223  name :  model.decoder.layers.3.encoder_attn.q_proj.weight\n","Count :  224  name :  model.decoder.layers.3.encoder_attn.q_proj.bias\n","Count :  225  name :  model.decoder.layers.3.encoder_attn.out_proj.weight\n","Count :  226  name :  model.decoder.layers.3.encoder_attn.out_proj.bias\n","Count :  227  name :  model.decoder.layers.3.encoder_attn_layer_norm.weight\n","Count :  228  name :  model.decoder.layers.3.encoder_attn_layer_norm.bias\n","Count :  229  name :  model.decoder.layers.3.fc1.weight\n","Count :  230  name :  model.decoder.layers.3.fc1.bias\n","Count :  231  name :  model.decoder.layers.3.fc2.weight\n","Count :  232  name :  model.decoder.layers.3.fc2.bias\n","Count :  233  name :  model.decoder.layers.3.final_layer_norm.weight\n","Count :  234  name :  model.decoder.layers.3.final_layer_norm.bias\n","Count :  235  name :  model.decoder.layers.4.self_attn.k_proj.weight\n","Count :  236  name :  model.decoder.layers.4.self_attn.k_proj.bias\n","Count :  237  name :  model.decoder.layers.4.self_attn.v_proj.weight\n","Count :  238  name :  model.decoder.layers.4.self_attn.v_proj.bias\n","Count :  239  name :  model.decoder.layers.4.self_attn.q_proj.weight\n","Count :  240  name :  model.decoder.layers.4.self_attn.q_proj.bias\n","Count :  241  name :  model.decoder.layers.4.self_attn.out_proj.weight\n","Count :  242  name :  model.decoder.layers.4.self_attn.out_proj.bias\n","Count :  243  name :  model.decoder.layers.4.self_attn_layer_norm.weight\n","Count :  244  name :  model.decoder.layers.4.self_attn_layer_norm.bias\n","Count :  245  name :  model.decoder.layers.4.encoder_attn.k_proj.weight\n","Count :  246  name :  model.decoder.layers.4.encoder_attn.k_proj.bias\n","Count :  247  name :  model.decoder.layers.4.encoder_attn.v_proj.weight\n","Count :  248  name :  model.decoder.layers.4.encoder_attn.v_proj.bias\n","Count :  249  name :  model.decoder.layers.4.encoder_attn.q_proj.weight\n","Count :  250  name :  model.decoder.layers.4.encoder_attn.q_proj.bias\n","Count :  251  name :  model.decoder.layers.4.encoder_attn.out_proj.weight\n","Count :  252  name :  model.decoder.layers.4.encoder_attn.out_proj.bias\n","Count :  253  name :  model.decoder.layers.4.encoder_attn_layer_norm.weight\n","Count :  254  name :  model.decoder.layers.4.encoder_attn_layer_norm.bias\n","Count :  255  name :  model.decoder.layers.4.fc1.weight\n","Count :  256  name :  model.decoder.layers.4.fc1.bias\n","Count :  257  name :  model.decoder.layers.4.fc2.weight\n","Count :  258  name :  model.decoder.layers.4.fc2.bias\n","Count :  259  name :  model.decoder.layers.4.final_layer_norm.weight\n","Count :  260  name :  model.decoder.layers.4.final_layer_norm.bias\n","Count :  261  name :  model.decoder.layers.5.self_attn.k_proj.weight\n","Count :  262  name :  model.decoder.layers.5.self_attn.k_proj.bias\n","Count :  263  name :  model.decoder.layers.5.self_attn.v_proj.weight\n","Count :  264  name :  model.decoder.layers.5.self_attn.v_proj.bias\n","Count :  265  name :  model.decoder.layers.5.self_attn.q_proj.weight\n","Count :  266  name :  model.decoder.layers.5.self_attn.q_proj.bias\n","Count :  267  name :  model.decoder.layers.5.self_attn.out_proj.weight\n","Count :  268  name :  model.decoder.layers.5.self_attn.out_proj.bias\n","Count :  269  name :  model.decoder.layers.5.self_attn_layer_norm.weight\n","Count :  270  name :  model.decoder.layers.5.self_attn_layer_norm.bias\n","Count :  271  name :  model.decoder.layers.5.encoder_attn.k_proj.weight\n","Count :  272  name :  model.decoder.layers.5.encoder_attn.k_proj.bias\n","Count :  273  name :  model.decoder.layers.5.encoder_attn.v_proj.weight\n","Count :  274  name :  model.decoder.layers.5.encoder_attn.v_proj.bias\n","Count :  275  name :  model.decoder.layers.5.encoder_attn.q_proj.weight\n","Count :  276  name :  model.decoder.layers.5.encoder_attn.q_proj.bias\n","Count :  277  name :  model.decoder.layers.5.encoder_attn.out_proj.weight\n","Count :  278  name :  model.decoder.layers.5.encoder_attn.out_proj.bias\n","Count :  279  name :  model.decoder.layers.5.encoder_attn_layer_norm.weight\n","Count :  280  name :  model.decoder.layers.5.encoder_attn_layer_norm.bias\n","Count :  281  name :  model.decoder.layers.5.fc1.weight\n","Count :  282  name :  model.decoder.layers.5.fc1.bias\n","Count :  283  name :  model.decoder.layers.5.fc2.weight\n","Count :  284  name :  model.decoder.layers.5.fc2.bias\n","Count :  285  name :  model.decoder.layers.5.final_layer_norm.weight\n","Count :  286  name :  model.decoder.layers.5.final_layer_norm.bias\n","Count :  287  name :  model.decoder.layernorm_embedding.weight\n","Count :  288  name :  model.decoder.layernorm_embedding.bias\n","Count :  289  name :  classification_head.dense.weight\n","Count :  290  name :  classification_head.dense.bias\n","Count :  291  name :  classification_head.out_proj.weight\n","Count :  292  name :  classification_head.out_proj.bias\n"]}],"source":["cnt = 0\n","for name, param in model_intent_only.named_parameters():\n","    print(\"Count : \", cnt, \" name : \", name)\n","    cnt+=1"]},{"cell_type":"code","execution_count":79,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":7,"status":"ok","timestamp":1684514987550,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"AO6_5sxbSpMl","outputId":"8f315432-1df1-4817-bfa2-f10deba8f7e7"},"outputs":[{"output_type":"stream","name":"stdout","text":["Total trainanable parameters :  0.0\n"]}],"source":["cnt = 0\n","for name, param in model_intent_only.named_parameters():\n","\n","    param.requires_grad = False    \n","    cnt+=1\n","        \n","\n","num_param = sum(p.numel() for p in model_intent_only.parameters() if p.requires_grad)\n","print(\"Total trainanable parameters : \", num_param/1e6)"]},{"cell_type":"code","source":["num_param_intent_shared = sum(p.numel() for p in model_intent_shared.parameters() if p.requires_grad)\n","print(num_param_intent_shared)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"6yeFoNpD_GOo","executionInfo":{"status":"ok","timestamp":1684514987551,"user_tz":-330,"elapsed":4,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"35227a55-c719-422d-ab44-0a5c9b520b4a"},"execution_count":80,"outputs":[{"output_type":"stream","name":"stdout","text":["1189644\n"]}]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/data/train_audio_data.p', 'rb') as f:\n","  train_audio_data = pickle.load(f)\n"],"metadata":{"id":"o8EXsSFqrNC8","executionInfo":{"status":"ok","timestamp":1684514992122,"user_tz":-330,"elapsed":4573,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":81,"outputs":[]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/data/valid_audio_data.p', 'rb') as f:\n","  valid_audio_data = pickle.load(f)\n"],"metadata":{"id":"Wm5X37KUrRRz","executionInfo":{"status":"ok","timestamp":1684514993289,"user_tz":-330,"elapsed":1168,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":82,"outputs":[]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/data/test_audio_data.p', 'rb') as f:\n","  test_audio_data = pickle.load(f)"],"metadata":{"id":"hw5LmU8urTE3","executionInfo":{"status":"ok","timestamp":1684514993866,"user_tz":-330,"elapsed":578,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":83,"outputs":[]},{"cell_type":"code","source":[],"metadata":{"id":"xoq3xwx0sPpV","executionInfo":{"status":"ok","timestamp":1684514993867,"user_tz":-330,"elapsed":9,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":83,"outputs":[]},{"cell_type":"code","source":["ACOUSTIC_DIM"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"bqMI2YA-sR6a","executionInfo":{"status":"ok","timestamp":1684514993867,"user_tz":-330,"elapsed":9,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"ac666c7f-98bb-4362-b5b6-925ade91e850"},"execution_count":84,"outputs":[{"output_type":"execute_result","data":{"text/plain":["768"]},"metadata":{},"execution_count":84}]},{"cell_type":"code","source":[],"metadata":{"id":"SjSkyKOcsTUR","executionInfo":{"status":"ok","timestamp":1684514993867,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":84,"outputs":[]},{"cell_type":"code","source":["len(train_audio_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"ljr5LwgArVHH","executionInfo":{"status":"ok","timestamp":1684514993867,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"8629f2ca-3a50-4573-a5ae-cbe9671082d8"},"execution_count":85,"outputs":[{"output_type":"execute_result","data":{"text/plain":["540"]},"metadata":{},"execution_count":85}]},{"cell_type":"code","source":["len(valid_audio_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"vDoX00-jrWwt","executionInfo":{"status":"ok","timestamp":1684514993867,"user_tz":-330,"elapsed":6,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"62be8dd8-6754-45ab-f2be-c49ebfafab1a"},"execution_count":86,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":86}]},{"cell_type":"code","source":["len(test_audio_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"5yYlZMY9rYXX","executionInfo":{"status":"ok","timestamp":1684514993867,"user_tz":-330,"elapsed":4,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"39a6b88b-c04b-4073-ed5e-cdffc6add45b"},"execution_count":87,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":87}]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/data/train_video_data.p', 'rb') as f:\n","  train_video_data = pickle.load(f)"],"metadata":{"id":"mtrKUc7FraUG","executionInfo":{"status":"ok","timestamp":1684515000962,"user_tz":-330,"elapsed":7097,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":88,"outputs":[]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/data/valid_video_data.p', 'rb') as f:\n","  valid_video_data = pickle.load(f)"],"metadata":{"id":"UNRLgS6GrbkD","executionInfo":{"status":"ok","timestamp":1684515001773,"user_tz":-330,"elapsed":813,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":89,"outputs":[]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/data/test_video_data.p', 'rb') as f:\n","  test_video_data = pickle.load(f)"],"metadata":{"id":"QISZcizPrdPI","executionInfo":{"status":"ok","timestamp":1684515001773,"user_tz":-330,"elapsed":9,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":90,"outputs":[]},{"cell_type":"code","source":["len(train_video_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Ky_y6XNerhbc","executionInfo":{"status":"ok","timestamp":1684515001774,"user_tz":-330,"elapsed":9,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"0fd77e3e-8ac5-4240-84aa-2a2b3190416c"},"execution_count":91,"outputs":[{"output_type":"execute_result","data":{"text/plain":["540"]},"metadata":{},"execution_count":91}]},{"cell_type":"code","source":["len(valid_video_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"4QPBizJRriqv","executionInfo":{"status":"ok","timestamp":1684515001774,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"898907e4-f70b-4293-b324-10cb21103bbf"},"execution_count":92,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":92}]},{"cell_type":"code","source":["len(test_video_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"S88jHfKPrkEq","executionInfo":{"status":"ok","timestamp":1684515001774,"user_tz":-330,"elapsed":6,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"b3610ba5-69c6-478e-c8a1-b66f72ec7d4d"},"execution_count":93,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":93}]},{"cell_type":"code","source":["ACOUSTIC_DIM"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"wFaAGgarsnM6","executionInfo":{"status":"ok","timestamp":1684515001774,"user_tz":-330,"elapsed":4,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"66bc890b-5cd1-4cb2-e9b2-32a8dcccabf0"},"execution_count":94,"outputs":[{"output_type":"execute_result","data":{"text/plain":["768"]},"metadata":{},"execution_count":94}]},{"cell_type":"code","source":["def pad_seq(tensor, dim, max_len):\n","  if max_len > tensor.shape[0] :\n","    return torch.cat([tensor, torch.zeros(max_len - tensor.shape[0], dim)])\n","  else:\n","    return tensor[:max_len]  "],"metadata":{"id":"JBpRVICisdoK","executionInfo":{"status":"ok","timestamp":1684515001774,"user_tz":-330,"elapsed":3,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":95,"outputs":[]},{"cell_type":"code","source":["train_audio_data = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","                                                   dim = ACOUSTIC_DIM,\n","                                                   max_len = ACOUSTIC_MAX_LEN)\n","                                                  for a in train_audio_data], 0)\n","train_audio_data.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"ySuA3QbssXIj","executionInfo":{"status":"ok","timestamp":1684515003518,"user_tz":-330,"elapsed":1747,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"a75444af-250a-4bd3-d5c1-4be185e57d36"},"execution_count":96,"outputs":[{"output_type":"execute_result","data":{"text/plain":["torch.Size([540, 1000, 768])"]},"metadata":{},"execution_count":96}]},{"cell_type":"code","source":["valid_audio_data = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","                                                   dim = ACOUSTIC_DIM,\n","                                                   max_len = ACOUSTIC_MAX_LEN)\n","                                                  for a in valid_audio_data], 0)\n","valid_audio_data.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"1E29NjJ3sgqE","executionInfo":{"status":"ok","timestamp":1684515003518,"user_tz":-330,"elapsed":9,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"fe031986-3daa-4e66-eb7b-3198cc67cbf7"},"execution_count":97,"outputs":[{"output_type":"execute_result","data":{"text/plain":["torch.Size([75, 1000, 768])"]},"metadata":{},"execution_count":97}]},{"cell_type":"code","source":["test_audio_data = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","                                                   dim = ACOUSTIC_DIM,\n","                                                   max_len = ACOUSTIC_MAX_LEN)\n","                                                  for a in test_audio_data], 0)\n","test_audio_data.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"lrP-ULZIsZk0","executionInfo":{"status":"ok","timestamp":1684515003518,"user_tz":-330,"elapsed":7,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"01789bae-f25b-4945-b9c1-6064d9673e10"},"execution_count":98,"outputs":[{"output_type":"execute_result","data":{"text/plain":["torch.Size([75, 1000, 768])"]},"metadata":{},"execution_count":98}]},{"cell_type":"code","source":["VISUAL_DIM"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"eBbHubGQsqNU","executionInfo":{"status":"ok","timestamp":1684515003518,"user_tz":-330,"elapsed":6,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"1cb87d1f-f124-4488-f78d-138f9dabecc5"},"execution_count":99,"outputs":[{"output_type":"execute_result","data":{"text/plain":["2048"]},"metadata":{},"execution_count":99}]},{"cell_type":"code","source":["VISUAL_MAX_LEN"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"PdLYvVIQsrOP","executionInfo":{"status":"ok","timestamp":1684515003519,"user_tz":-330,"elapsed":6,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"d8394151-dce2-4745-91c9-5e99dc835511"},"execution_count":100,"outputs":[{"output_type":"execute_result","data":{"text/plain":["480"]},"metadata":{},"execution_count":100}]},{"cell_type":"code","source":["train_video_data = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","                                                   dim = VISUAL_DIM,\n","                                                   max_len = VISUAL_MAX_LEN)\n","                                                  for a in train_video_data], 0)\n","train_video_data.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"YqS2rHvussrR","executionInfo":{"status":"ok","timestamp":1684515003519,"user_tz":-330,"elapsed":4,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"0af995a4-31f6-40af-f7cb-f535e4e4aa8c"},"execution_count":101,"outputs":[{"output_type":"execute_result","data":{"text/plain":["torch.Size([540, 480, 2048])"]},"metadata":{},"execution_count":101}]},{"cell_type":"code","source":["valid_video_data = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","                                                   dim = VISUAL_DIM,\n","                                                   max_len = VISUAL_MAX_LEN)\n","                                                  for a in valid_video_data], 0)\n","valid_video_data.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"2M9vVoOLst5m","executionInfo":{"status":"ok","timestamp":1684515004320,"user_tz":-330,"elapsed":804,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"bacf8822-45bb-4647-8c30-2f6bb705b56c"},"execution_count":102,"outputs":[{"output_type":"execute_result","data":{"text/plain":["torch.Size([75, 480, 2048])"]},"metadata":{},"execution_count":102}]},{"cell_type":"code","source":["test_video_data = torch.stack([pad_seq(torch.tensor(a, dtype = torch.float),\n","                                                   dim = VISUAL_DIM,\n","                                                   max_len = VISUAL_MAX_LEN)\n","                                                  for a in test_video_data], 0)\n","test_video_data.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"zd-V7JFasvBb","executionInfo":{"status":"ok","timestamp":1684515004320,"user_tz":-330,"elapsed":12,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"c4ca5b30-df87-4fcd-86ac-b37559d88943"},"execution_count":103,"outputs":[{"output_type":"execute_result","data":{"text/plain":["torch.Size([75, 480, 2048])"]},"metadata":{},"execution_count":103}]},{"cell_type":"code","source":["with open(\"/content/drive/MyDrive/Colab Notebooks/35/data/train_text_data.p\", \"rb\") as f:\n","    train_text_data = pickle.load(f)\n","\n","len(train_text_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"v62NuYK7szPA","executionInfo":{"status":"ok","timestamp":1684515004321,"user_tz":-330,"elapsed":10,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"9bffe6ee-b23d-4d0d-86fc-82862a619a1f"},"execution_count":104,"outputs":[{"output_type":"execute_result","data":{"text/plain":["540"]},"metadata":{},"execution_count":104}]},{"cell_type":"code","source":["with open(\"/content/drive/MyDrive/Colab Notebooks/35/data/valid_text_data.p\", \"rb\") as f:\n","    valid_text_data = pickle.load(f)\n","\n","len(valid_text_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"uUZjgBD8s0ka","executionInfo":{"status":"ok","timestamp":1684515004321,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"a1393c7c-acec-4943-a86d-78f3994f715b"},"execution_count":105,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":105}]},{"cell_type":"code","source":["with open(\"/content/drive/MyDrive/Colab Notebooks/35/data/test_text_data.p\", \"rb\") as f:\n","    test_text_data = pickle.load(f)\n","\n","len(test_text_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"V6Pa7Hhfs128","executionInfo":{"status":"ok","timestamp":1684515004321,"user_tz":-330,"elapsed":6,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"a9c783c7-1f2a-4fb1-95aa-41be0a7a6bde"},"execution_count":106,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":106}]},{"cell_type":"code","source":["with open(\"/content/drive/MyDrive/Colab Notebooks/35/data/train_label_data.p\", \"rb\") as f:\n","    train_label_data = pickle.load(f)\n","\n","len(train_label_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"VQLYtXdls3SZ","executionInfo":{"status":"ok","timestamp":1684515005525,"user_tz":-330,"elapsed":1207,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"321a8602-e0e8-4702-cf7f-3158300f9b39"},"execution_count":107,"outputs":[{"output_type":"execute_result","data":{"text/plain":["540"]},"metadata":{},"execution_count":107}]},{"cell_type":"code","source":["with open(\"/content/drive/MyDrive/Colab Notebooks/35/data/valid_label_data.p\", \"rb\") as f:\n","    valid_label_data = pickle.load(f)\n","\n","len(valid_label_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"32-Y_p7Ns4ry","executionInfo":{"status":"ok","timestamp":1684515005526,"user_tz":-330,"elapsed":18,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"f98996c1-196d-48a0-8e86-d384a753ce05"},"execution_count":108,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":108}]},{"cell_type":"code","source":["with open(\"/content/drive/MyDrive/Colab Notebooks/35/data/test_label_data.p\", \"rb\") as f:\n","    test_label_data = pickle.load(f)\n","\n","len(test_label_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"0JFvqWkms6LR","executionInfo":{"status":"ok","timestamp":1684515005526,"user_tz":-330,"elapsed":14,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"4b5a4070-3880-4710-f0b4-6a509e17d486"},"execution_count":109,"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":109}]},{"cell_type":"code","execution_count":110,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":13,"status":"ok","timestamp":1684515005527,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"9Ik_vBo7x6Rm","outputId":"24f0fc63-9d80-4a37-b9db-e5e1478a028f"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["BartTokenizerFast(name_or_path='facebook/bart-base', vocab_size=50265, model_max_length=1024, is_fast=True, padding_side='right', truncation_side='right', special_tokens={'bos_token': '<s>', 'eos_token': '</s>', 'unk_token': '<unk>', 'sep_token': '</s>', 'pad_token': '<pad>', 'cls_token': '<s>', 'mask_token': AddedToken(\"<mask>\", rstrip=False, lstrip=True, single_word=False, normalized=False)}, clean_up_tokenization_spaces=True)"]},"metadata":{},"execution_count":110}],"source":["tokenizer"]},{"cell_type":"code","execution_count":111,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":10,"status":"ok","timestamp":1684515005527,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"X0y5DYklxuUT","outputId":"ba764875-e092-408f-ce85-8073975ba8ff"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["BartTokenizerFast(name_or_path='facebook/bart-base', vocab_size=50265, model_max_length=1024, is_fast=True, padding_side='right', truncation_side='right', special_tokens={'bos_token': '<s>', 'eos_token': '</s>', 'unk_token': '<unk>', 'sep_token': '</s>', 'pad_token': '<pad>', 'cls_token': '<s>', 'mask_token': AddedToken(\"<mask>\", rstrip=False, lstrip=True, single_word=False, normalized=False), 'additional_special_tokens': ['[CONTEXT]', '[UTTERANCE]']}, clean_up_tokenization_spaces=True)"]},"metadata":{},"execution_count":111}],"source":["p = {\n","        'additional_special_tokens' : ['[CONTEXT]', '[UTTERANCE]']\n","    }\n","\n","tokenizer.add_special_tokens(p)\n","\n","tokenizer"]},{"cell_type":"code","execution_count":112,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":7,"status":"ok","timestamp":1684515005527,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"P1Ctmwokx-T2","outputId":"4496ecde-9489-400a-bfb0-04ca5618f348"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["Embedding(50267, 768)"]},"metadata":{},"execution_count":112}],"source":["model_sarcasm_only.resize_token_embeddings(len(tokenizer))"]},{"cell_type":"code","execution_count":113,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":1431,"status":"ok","timestamp":1684515006954,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"Zzg9pjhb9flX","outputId":"af8d6f2b-10aa-48bd-db3f-8728a4a8c35a"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["Embedding(50267, 768)"]},"metadata":{},"execution_count":113}],"source":["model_intent_only.resize_token_embeddings(len(tokenizer))"]},{"cell_type":"code","execution_count":114,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":83,"status":"ok","timestamp":1684515006954,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"SY3KgbQUKoFd","outputId":"88ec8a7e-7376-4531-c053-0e4938d936de"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["['<s>', '</s>', '<unk>', '<pad>', '<mask>', '[CONTEXT]', '[UTTERANCE]']"]},"metadata":{},"execution_count":114}],"source":["tokenizer.all_special_tokens"]},{"cell_type":"code","execution_count":115,"metadata":{"id":"M7uPY9VkLRcr","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":82,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["def prepare_dataset(text_data):\n","                    \n","\n","            dialog = []\n","            labels = []\n","            sample_utterances = []\n","            for i in range(len(text_data)):\n","                data_point = text_data[i]\n","\n","                example_speaker = data_point['speaker']\n","                example_utterance = data_point['utterance']\n","                temp_label = int(data_point['sarcasm'])\n","                sample_utterances.append(data_point)\n","\n","                # example_dialog = '[CONTEXT] '\n","                # example_dialog = '[TARGET] '\n","                example_dialog = '[CONTEXT] '\n","\n","\n","                for speaker, utterance in list(zip(data_point['context_speakers'], data_point['context'])):\n","                    example_dialog = example_dialog + speaker.upper() + \" : \" + utterance + \" | \"\n","\n","                example_dialog = example_dialog + ' [UTTERANCE] ' + example_speaker + \" : \" + example_utterance + \" | \"\n","                # example_dialog = example_dialog + example_speaker + \" : \" + example_utterance\n","                # example_dialog = example_dialog + example_speaker + \" : \" + example_utterance \n","                # print(example_dialog)\n","                example_dialog = re.sub(' +', ' ', example_dialog)\n","\n","                dialog.append(example_dialog)\n","                labels.append(temp_label)\n","\n","                # print(data_point)\n","                # print(temp_label)\n","\n","            # df = pd.DataFrame(dialog, columns=['dialog'])\n","\n","            labels = torch.tensor(labels, dtype=torch.long)\n","\n","            \n","\n","            enc = tokenizer(dialog, max_length = SOURCE_MAX_LEN, padding = 'max_length', truncation = True)\n","\n","            # df['audio_features'] = acoustic_data\n","            # df['visual_features'] = visual_data\n","\n","            return torch.tensor(enc['input_ids'], dtype=torch.long), torch.tensor(enc['attention_mask'], dtype=torch.bool)\n"]},{"cell_type":"code","source":["len(train_text_data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"mg32FNK5tFoH","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":82,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"ba351404-aabd-40b1-a9fc-58d3e07b55a9"},"execution_count":116,"outputs":[{"output_type":"execute_result","data":{"text/plain":["540"]},"metadata":{},"execution_count":116}]},{"cell_type":"code","source":["train_text_data"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"AzEFchUtxKm0","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":81,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"f8a79f91-a751-4530-9ebd-acc00addfa1e"},"execution_count":117,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[{'utterance': \"It's just a privilege to watch your mind at work.\",\n","  'speaker': 'SHELDON',\n","  'context': ['I never would have identified the fingerprints of string theory in the aftermath of the Big Bang.',\n","   \"My apologies. What's your plan?\"],\n","  'context_speakers': ['LEONARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"If you're compiling a mix CD for a double suicide. Oh, I hope that scratching post is for you.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Oh, good Lord.', \"God, that's a good song.\"],\n","  'context_speakers': ['SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': '- Now?! - No, after my tongue has swollen to the size of a brisket!',\n","  'speaker': 'HOWARD',\n","  'context': ['Take that!',\n","   \"What's the matter?\",\n","   'This granola bar has peanuts in it!',\n","   'Oh, my God. Why did you eat it?',\n","   \"I don't know. It was just there!\",\n","   'If I had a gun there, would you have shot yourself?',\n","   \"Don't yell at me! I've got to go to the emergency room!\"],\n","  'context_speakers': ['LEONARD',\n","   'LEONARD',\n","   'HOWARD',\n","   'LEONARD',\n","   'HOWARD',\n","   'LEONARD',\n","   'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Notify the editors of the Oxford English Dictionary. The word \"plenty\" has been redefined to mean \"two.\"',\n","  'speaker': 'SHELDON',\n","  'context': [\"Are you saying that you've been holding back?\",\n","   'Of course. Out of respect.',\n","   'So, how do you explain the ten years before Penny?',\n","   'Who were you respecting then?',\n","   \"Well, I've dated plenty of women.\",\n","   'There was Joyce Kim, Leslie Winkle'],\n","  'context_speakers': ['HOWARD',\n","   'LEONARD',\n","   'HOWARD',\n","   'RAJ',\n","   'LEONARD',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'They gave him an assistant? If I want a new pen, I have to go to the bank with wire cutters.',\n","  'speaker': 'HOWARD',\n","  'context': [\"They're going to digitally add a supernova.\\nThey say it's the perfect metaphor for my incandescent talent.\",\n","   'Right, a ball of hot, flaming gas that collapses upon itself.',\n","   \"Excuse me. Oh, it's my assistant, Trevor.\",\n","   'Go for Koothrappali.'],\n","  'context_speakers': ['RAJ', 'SHELDON', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Maybe when I publish it, I'll include an inhaler.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"Excuse me, Dr. Cooper. I'm Kathy O'Brien.\",\n","   'I just finished reading your paper reconciling the black hole information paradox with your theory of string-network condensates, and it just took my breath away.'],\n","  'context_speakers': ['PERSON', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'The blunt instrument that will be the focus of my murder trial?',\n","  'speaker': 'LEONARD',\n","  'context': ['What, Sheldon? What, Sheldon?! What, Sheldon?!',\n","   'Tell me what you see here.'],\n","  'context_speakers': ['LEONARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': '\"Dave.\" Sounds like Leonard\\'s got a new BFF.',\n","  'speaker': 'HOWARD',\n","  'context': ['There are always do-overs when my people play sports.',\n","   \"Where were you that's more important than Wii bowling night?\",\n","   'Actually...',\n","   \"It's rhetorical.\\nThere is nothing more important than Wii bowling night.\",\n","   \"Come on, it's just a video game. And we suck at it.\",\n","   'Nice motivational speech from the team captain.',\n","   'Where were you?'],\n","  'context_speakers': ['HOWARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Why don't you just lie down on the floor and swim there?\",\n","  'speaker': 'LEONARD',\n","  'context': ['Where are we going?',\n","   \"You're driving me to the mall.\",\n","   \"I'm going to acquire a book that summarizes the current theories in the field of friend-making.\"],\n","  'context_speakers': ['LEONARD', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I hope I'm a waitress at the Cheesecake Factory for my whole life.\",\n","  'speaker': 'PENNY',\n","  'context': ['Ask me why.',\n","   'Do I have to?',\n","   \"Of course. That's how you move a conversation forward.\",\n","   'Why?',\n","   'The learning opportunities would be abundant.',\n","   'Additionally, I like having my belly scratched.'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'You think \"coffee,\" means coffee. That is so sweet.',\n","  'speaker': 'PENNY',\n","  'context': [\"Come on. I think it's nice that Captain Sweatpants showed up to your art opening.\",\n","   \"Yeah, it would have been nicer if he hadn't touched all the cheese.\",\n","   \"You know, it's kind of early.\\nDo you want to maybe come in for some coffee?\",\n","   \"Gee, it's a little late for coffee, isn't it?\"],\n","  'context_speakers': ['PENNY', 'PERSON', 'PENNY', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"It's an iCal download. She can put it right in her phone.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Alright, want some French toast?',\n","   \"It's oatmeal day.\",\n","   \"Tell you what, next French toast day, I'll make you oatmeal.\",\n","   'Dear Lord, are you still going to be here on French toast day?',\n","   'Morning.',\n","   'Look, Leonard, Penny made French toast.',\n","   \"Sorry, I haven't given her your schedule yet.\"],\n","  'context_speakers': ['PENNY',\n","   'SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"You know, I've always wanted to go to a Goth nightclub.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Howard, what did you do?',\n","   \"They're called tattoo sleeves. Look.\",\n","   'I got them online. Raj got a set, too.',\n","   'Fantastic, right?',\n","   'Put them on, have hot sex with some freaky girl with her business pierced, take them off, and I can still be buried in a Jewish cemetery.'],\n","  'context_speakers': ['LEONARD', 'HOWARD', 'HOWARD', 'HOWARD', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, we don't need to talk about that anymore. Leonard covered it pretty well in the car.\",\n","  'speaker': 'PENNY',\n","  'context': ['So, what should we talk about?',\n","   'Well, we could always learn more about why people who believe in psychics are idiots. Leonard?'],\n","  'context_speakers': ['BERNADETTE', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"There a new girlfriend in there? 'Cause you might need one.\",\n","  'speaker': 'PENNY',\n","  'context': ['Leonard, what do the two of you talk about after the coitus?',\n","   'My guess is, \"Hey, four minutes! New record!\"',\n","   \"That's why I'm the funny one.\",\n","   'Anyway, we followed the guy to this garage sale, and they had the coolest stuff.',\n","   'They were closing up. We got this whole box for 60 bucks.',\n","   \"We didn't even get to go through it all. There could be anything in here.\"],\n","  'context_speakers': ['SHELDON',\n","   'HOWARD',\n","   'HOWARD',\n","   'LEONARD',\n","   'HOWARD',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'We might be starting to zero in on your problem.',\n","  'speaker': 'LEONARD',\n","  'context': ['Are you guys doing okay?',\n","   'Leonard and I are fine, but I think Raj needs to meet a girl really soon.',\n","   \"Well, that shouldn't be too hard. He's such a cutie.\",\n","   'Thank you, but \"cute\" is for bunnies. I want to be something with sex appeal, like... like a labradoodle.\\nLabradoodle?'],\n","  'context_speakers': ['BERNADETTE', 'HOWARD', 'BERNADETTE', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"That's got to be fun for you, huh? Watching Hofstadter suck the saliva out of your sister's mouth?\",\n","  'speaker': 'HOWARD',\n","  'context': ['The audible sigh is a show of exasperation, right?',\n","   'Right.',\n","   'Is it about me or the dead pixels?',\n","   \"Hey, you're just in time. We made Tex-Mex.\",\n","   'Oh, sounds great.',\n","   'Oh, Indians making Tex-Mex.',\n","   'Might as well have had the Chinese pizza.',\n","   'Hey, you.'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'RAJ',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'He drank from Leonard\\'s glass. \"He drank from Leonard\\'s glass\"-- the words they\\'ll be carving into my tombstone.',\n","  'speaker': 'SHELDON',\n","  'context': ['of your subtropical girlfriend.',\n","   \"Hey! That's my sister and my country you're talking about.\",\n","   \"Leonard may have defiled one, but I won't have you talking smack about the other.\",\n","   'You guys ready to order?',\n","   \"Yes, I'd like a seven-day course of penicillin, some, uh, syrup of ipecac-- to induce vomiting-- and a mint.\",\n","   \"I don't understand.\"],\n","  'context_speakers': ['SHELDON',\n","   'RAJ',\n","   'RAJ',\n","   'BERNADETTE',\n","   'SHELDON',\n","   'BERNADETTE'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"is an event that's like a heart attack.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"Yeah, I'm saying, it'd be easier to lift a car.\",\n","   'What can I tell you? After I found the courage to put her pants back on, I was unstoppable.',\n","   'So, how is she?',\n","   \"They're running tests-- I don't know.\",\n","   'It may have been a heart attack or heart-attack-like event.',\n","   \"What's the difference?\",\n","   'A heart-attack-like event'],\n","  'context_speakers': ['PENNY',\n","   'HOWARD',\n","   'LEONARD',\n","   'HOWARD',\n","   'HOWARD',\n","   'PENNY',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Congratulations and welcome temporarily aboard. Here's your I.D. card, your key and your lapel pin. Which Leonard was too cool to wear.\",\n","  'speaker': 'SHELDON',\n","  'context': ['And a living will and durable power of attorney.',\n","   'This says you can make \"end of life\" decisions for me.',\n","   \"As your friend, let's hope it doesn't come to that. \\nPlease sign.\",\n","   'Did you sign this?',\n","   \"There's a reciprocity clause.\\nYou get to pull the plug on him, too.\",\n","   'Well, that seems fair.'],\n","  'context_speakers': ['SHELDON', 'RAJ', 'SHELDON', 'RAJ', 'LEONARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Oh, I could so see you being the face of hemorrhoids.',\n","  'speaker': 'AMY',\n","  'context': [\"It's like a waterfall of liquid gold.\",\n","   'My turn.',\n","   \"You know, I don't even know what the point of me staying in L.A. is.\",\n","   \"I haven't gotten a single acting job since I moved out here.\",\n","   'The closest I came was last month, I got a callback for a hemorrhoid commercial.'],\n","  'context_speakers': ['AMY', 'AMY', 'PENNY', 'PENNY', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"When you're done copping a feel, that goes on my wrist.\",\n","  'speaker': 'AMY',\n","  'context': [\"Oh, sweetie, guests don't normally wear corsages to a wedding. \\nThat's more of a prom thing.\",\n","   'I never went to my prom. \\nMy mom paid my cousin to take me, but he just used the money to buy drugs.',\n","   'Put the corsage on her.',\n","   \"When you're done copping a feel, that goes on my wrist.\"],\n","  'context_speakers': ['PENNY', 'AMY', 'PENNY', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Smart. Whisper so the deaf chick can't hear you.\",\n","  'speaker': 'HOWARD',\n","  'context': ['Wow.',\n","   \"You're engaged to my friend.\",\n","   \"Hey, Bernadette doesn't mind where I get my motor running, as long as I park in the right garage.\",\n","   \"I can't believe you're engaged to my friend.\"],\n","  'context_speakers': ['HOWARD', 'PENNY', 'HOWARD', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"You just got off the list. Would you like back on it? This'll just take a moment. It's on a five and a quarter inch floppy.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"What differences does it make if Leonard goes to Wil Wheaton's party?\",\n","   \"Wil Wheaton is Sheldon's mortal enemy.\",\n","   'Mortal enemy?',\n","   \"Mm-hmm. Sheldon, I know you're a bit of a left-handed monkey wrench, but...\\nyou really have a mortal enemy?\",\n","   'In fact, I have 61 of them.',\n","   'Would you like to see the list?',\n","   'Oh, say no, say no, say no, say no.'],\n","  'context_speakers': ['AMY',\n","   'PENNY',\n","   'AMY',\n","   'AMY',\n","   'SHELDON',\n","   'SHELDON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Frankly, Mom, I'm encouraged to see how advanced your group has become-- willing to sail out into the ocean without fear of falling off the edge.\",\n","  'speaker': 'SHELDON',\n","  'context': ['All-you-can-eat Last Supper Buffet. \\nAnd my personal favorite, Gunning with God.',\n","   \"What's Gunning with God? I'm afraid to ask.\",\n","   'Oh, it is a hoot and a half.',\n","   \"You write your sins on a clay pigeon, they fire 'em up in the air, and you pulverize them with a 12-gauge shotgun full of our Lord's forgiveness.\"],\n","  'context_speakers': ['PERSON', 'LEONARD', 'PERSON', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, my God, that's where I know you from.\",\n","  'speaker': 'PENNY',\n","  'context': ['Yeah, sure, come on in.',\n","   'Thanks.',\n","   'Want some mac and cheese?',\n","   'No. Lactose. Gas.',\n","   'Glass of wine?',\n","   'No. Sulfites. Migraines.',\n","   \"Okay, well, I'd offer you Halloween candy, but that's gone.\",\n","   \"So, what's up?\",\n","   'Okay, we used to go out, right?'],\n","  'context_speakers': ['PENNY',\n","   'LEONARD',\n","   'PENNY',\n","   'LEONARD',\n","   'PENNY',\n","   'LEONARD',\n","   'PENNY',\n","   'PENNY',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Hmm. That's too bad. We could have spent New Year's Eve waiting for the ball to drop.\",\n","  'speaker': 'PENNY',\n","  'context': ['Oh, oh, oh.',\n","   'Was he the one who made you eat your arm hair?',\n","   \"No, but, actually, that was this guy's sister.\",\n","   'All right, well, what do you think he wants?',\n","   \"I don't know.\",\n","   'You know, the holidays are just around the corner.',\n","   'Maybe he wants to see if he can lodge the other testicle up there.',\n","   'I told you. That was a different guy.'],\n","  'context_speakers': ['SHELDON',\n","   'SHELDON',\n","   'LEONARD',\n","   'PENNY',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"You'd think a winner could make a decent cup of cocoa.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Then take away his keys and make him wander the streets with the other drunks.',\n","   'You remember Sheldon and Raj and Howard.',\n","   'Not really, no.',\n","   \"It's funny, huh, Leonard?\",\n","   \"Back in school, I was the winner and you were the loser.\\nAnd now we're reversed.\",\n","   \"You're the winner.\"],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'PERSON',\n","   'PERSON',\n","   'PERSON',\n","   'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Hope your hands are steady. It's the width of a single hair. But this is just biology, so I'm sure it's no problem for a genius like you.\",\n","  'speaker': 'AMY',\n","  'context': ['All right.',\n","   \"I'm no stranger to a little gray matter.\",\n","   'Locus coeruleus.',\n","   'Locus coeruleus.',\n","   \"You're getting warmer-- it is, indeed, in the brain.\"],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'SHELDON', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Darn. If you weren't busy, I'd ask you to join us.\",\n","  'speaker': 'SHELDON',\n","  'context': ['\"Release the Kraken.\" That never gets old.',\n","   '\"Release the Kraken!\" Oh, chills.',\n","   \"Besides, I'm having dinner with Amy.\",\n","   \"Oh, okay. I'll just go home and be alone.\",\n","   'W-Which is cool.\\nI eat alone, I sleep alone, I cry alone, so...\\nCool.'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I'm starting to get a sense of it.\",\n","  'speaker': 'LEONARD',\n","  'context': [\"My tongue won't go as far forward as it used to.\",\n","   'Maybe your tongue is shrinking.',\n","   'Nope.',\n","   'Oh, you have no idea how annoying this is.'],\n","  'context_speakers': ['SHELDON', 'LEONARD', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I didn't know they gave Nobel prizes for making boom-boom in the potty.\",\n","  'speaker': 'LEONARD',\n","  'context': ['the paper was rejected, but recently, he was proven right.',\n","   \"And now he's on the fast track to win a Nobel prize.\",\n","   \"Yeah, that's basically what I said.\",\n","   'The point is Higgs is being celebrated for work he did 50 years ago, so that got me thinking,',\n","   \"perhaps I've already hit upon the idea that will win me my Nobel prize.\"],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'PENNY', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"You're right, the party's fantastic. Please, tell me more. I haven't heard enough about it all week because hearing about that never gets old!\",\n","  'speaker': 'HOWARD',\n","  'context': ['It was a compromise. I lost.',\n","   \"Can you believe Stuart's walking around taking credit for this party?\",\n","   'Who cares?',\n","   'What do you mean \"who cares\"? Look at what I pulled off here.\\nThe deejay\\'s on fire, there\\'s a Tardis photo booth in the back room, and, oh, my God, the food!',\n","   'Stuart wanted Kraft Draculoni and cheese.'],\n","  'context_speakers': ['SHELDON', 'RAJ', 'HOWARD', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Exactly what it was half an hour ago. It's like you're not even trying to get better.\",\n","  'speaker': 'SHELDON',\n","  'context': ['(thermometer beeps) 102.2.\\n(coughs)'],\n","  'context_speakers': ['SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Why don't you ask Scotty to beam you up?\",\n","  'speaker': 'PERSON',\n","  'context': ['Hello.',\n","   'Uh, four glasses of water, please.',\n","   'Anything for you guys?',\n","   'Can I use your phone? \\nOur car got stolen.'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'I was gonna say we pick up a cake or a pie. But an insult to her faith is always thoughtful.',\n","  'speaker': 'HOWARD',\n","  'context': [\"We can't show up to your mom's empty-handed. \\nWe should bring something.\",\n","   \"I already am.\\nI'm bringing the gift of knowledge.\",\n","   'Oh, boy.',\n","   \"Despite what her Bible says, recent archeological studies indicate that although camels are referenced in the Old Testament, they didn't exist in the Middle East until hundreds of years later.\"],\n","  'context_speakers': ['HOWARD', 'SHELDON', 'HOWARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Well, my beer isn't flat and my rack's not saggy. So far, the future's great.\",\n","  'speaker': 'PENNY',\n","  'context': [\"You're all inside it, and we hurtle through the very fabric of time.\",\n","   '(electronic sound effects playing) And welcome to 20 years in the future.',\n","   'Hey, you guys just time-traveled.\\nStop looking so bored.'],\n","  'context_speakers': ['RAJ', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"And if it turns out you don't... - you and I could live together. - You and... Oh, sure.\",\n","  'speaker': 'AMY',\n","  'context': [\"Ugh! It's like walking into the Amazon.\\nAnd not the good Amazon with one-day shipping.\\nThe awful one with birds and snakes.\",\n","   'You hate the sound of all those keys on his key chain.',\n","   'Four keys!\\nWho does he think he is, a warden?',\n","   \"See? Maybe you'll love living alone.\",\n","   \"I don't know. Perhaps.\"],\n","  'context_speakers': ['SHELDON', 'AMY', 'SHELDON', 'AMY', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"There they go, fighting again. You'd never hear her talk that way to Sauvignon Blanc.\",\n","  'speaker': 'AMY',\n","  'context': ['Howard and Bernadette, Raj and his girlfriend, Penny and Chardonnay, Penny and you.',\n","   \"There isn't any ranking.\",\n","   \"And if there were, we wouldn't be at the bottom of it.\",\n","   'Yeah, and actually, I drink Sauvignon Blanc.',\n","   \"That's the part you have a problem with?\",\n","   'Relax.'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'LEONARD',\n","   'PENNY',\n","   'LEONARD',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"On the bright side, that Oxford study was right. One friend down. I wonder who you're going to lose next.\",\n","  'speaker': 'SHELDON',\n","  'context': ['You would leave me out of the conversation with the next guy if you dump Leonard?',\n","   'Why say it? Yes!',\n","   \"Well, I would never leave you off my list, and not just because, without you, we're playing fast and loose with the word list.\",\n","   'Well, good job. Now she hates me.',\n","   'Great.'],\n","  'context_speakers': ['RAJ', 'PENNY', 'RAJ', 'PENNY', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Boy, when you met Bernadette, the field of robotics really took a hit.',\n","  'speaker': 'RAJ',\n","  'context': ['Robot girlfriend.',\n","   'Mm, that was mine.',\n","   'Robot prostitute.',\n","   'Also mine.',\n","   \"Wait. I'm confused. \\nWhy would you need both a robot girlfriend and a robot prostitute?\",\n","   \"There's just some things you don't do with your robot girlfriend.\\nBoy, when you met Bernadette,\"],\n","  'context_speakers': ['LEONARD',\n","   'HOWARD',\n","   'LEONARD',\n","   'HOWARD',\n","   'SHELDON',\n","   'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Ah, that's a good question. Apparently someone was being awfully flirty while not wearing their engagement ring, causing another someone to show up here thinking the first someone might be available.\",\n","  'speaker': 'LEONARD',\n","  'context': ['A gallant man would defend his fiance for being called a urethra.',\n","   'Can we please change the subject?',\n","   \"Well, that didn't go your way.\",\n","   'Dr. Lorvis, what are you doing here?',\n","   'Uh, well, actually I came to see you.',\n","   'Really? Why?'],\n","  'context_speakers': ['HOWARD', 'LEONARD', 'RAJ', 'PENNY', 'PERSON', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I don't think I'll be able to stop thinking about it.\",\n","  'speaker': 'PENNY',\n","  'context': ['This is one of my favorite places to kick back after a quest. They have a great house ale.',\n","   'Cool tiger.',\n","   \"Yeah, I've had him since level ten.\",\n","   'His name is Buttons.',\n","   'Anyway, if you had your own game character we could hang out, maybe go on a quest.',\n","   'That sounds interesting.',\n","   \"You'll think about it?\"],\n","  'context_speakers': ['HOWARD',\n","   'PENNY',\n","   'HOWARD',\n","   'HOWARD',\n","   'HOWARD',\n","   'PENNY',\n","   'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Good idea, sit with her. Hold her, comfort her. And if the moment feels right, see if you can cop a feel.',\n","  'speaker': 'HOWARD',\n","  'context': [\"It's smashed beyond repair. What are you gonna do with it?\",\n","   'What else? Sell it on eBay as \"slightly used.\"',\n","   \"It was Penny's boyfriend's, they broke up.\",\n","   'Apparently, he posted intimate details of their physical relationship on his blog which I cannot find anywhere.',\n","   \"I'm gonna go back and try talking to her again.\"],\n","  'context_speakers': ['HOWARD', 'RAJ', 'LEONARD', 'HOWARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': '- Maybe I heard you wrong. A lot of words sound like \"yoga.\"',\n","  'speaker': 'LEONARD',\n","  'context': ['Leonard. Come, join us.',\n","   'Hey, Dave.\\nAnd Penny, what a surprise.',\n","   'Dave was just showing me around the university. This place is unbelievable!',\n","   \"I know, I've been offering to show you around for a year and a half. You always said you had yoga.\"],\n","  'context_speakers': ['PERSON', 'LEONARD', 'PENNY', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Hardly a day goes by when I don't think about it.\",\n","  'speaker': 'LEONARD',\n","  'context': [\"You went to President Seibert's house in the middle of the night?\",\n","   \"He didn't respond to my e-mail, his phone number is unlisted. Tell me what my other option was.\",\n","   'You could have waited until morning.',\n","   \"I know, look who I'm talking to.\",\n","   'Do you remember the grant proposal I submitted to the National Science Foundation to detect slow-moving monopoles at the magnetic North Pole?'],\n","  'context_speakers': ['LEONARD', 'SHELDON', 'LEONARD', 'LEONARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Sure, maybe a tableau of me trying to pummel you to death.',\n","  'speaker': 'LEONARD',\n","  'context': [\"Of course not. I'm a theoretical physicist.\\nA career I chose in no small part because it's indoors, but if I'm able to detect slow-moving magnetic monopoles there, I will be the scientis to confirm string theory.\",\n","   'People will write books about me.',\n","   'Third-graders will create macaroni-art dioramas depicting scenes from my life.'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I don't know. Somebody bigger and... Yeah, sure. Why not you?\",\n","  'speaker': 'PERSON',\n","  'context': ['Stupid Aquaman.',\n","   'If it makes you feel any better, when I was dating Penny, she used to flip out on me all the time.',\n","   'Whoa. You dated Penny?',\n","   \"She didn't tell you?\",\n","   'She told me she dated a guy named Leonard.',\n","   'Who would have thought it was you?',\n","   'Who else would it be?'],\n","  'context_speakers': ['RAJ',\n","   'LEONARD',\n","   'PERSON',\n","   'LEONARD',\n","   'PERSON',\n","   'PERSON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, yeah, yeah. It's real great. Today I drove to Van Nuys for an audition I thought was going to be for a cat food commercial. Turned out to be porn.\",\n","  'speaker': 'PENNY',\n","  'context': [\"Thank you, Leonard. That doesn't make it sound boring at all.\",\n","   'So, how you holding up?',\n","   \"I'm fine.\",\n","   \"Oh, who are you kidding? She's breathtaking.\",\n","   \"So, Penny, Leonard tells me you're an actress. That must be pretty exciting.\"],\n","  'context_speakers': ['PERSON', 'AMY', 'PENNY', 'AMY', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Yes, there's nothing quite like the slightly widened eyes of mildly startled.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Okay, all right.',\n","   'That one was clever.',\n","   'Skeleton with phosphorous on a zip line.',\n","   'Come on out, merry pranksters.',\n","   'Take a bow.',\n","   \"You should've seen your face.\"],\n","  'context_speakers': ['SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, of course he is. She's very interesting. Did you know, when she was 14, she severed the webbing between her own toes?\",\n","  'speaker': 'SHELDON',\n","  'context': ['consider changing disciplines.\\nYeah, to the humanities-- perhaps history.',\n","   \"One of the advantages of teaching history is that you don't have to create things.\",\n","   'You know, you just have to remember stuff that happened and then parrot it back.',\n","   'You could have fun with that.',\n","   \"Yeah, that's not it.\",\n","   \"Stuart's kind of interested in Amy.\"],\n","  'context_speakers': ['SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'LEONARD',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Yeah, I'm really looking forward to it. In fact, there's no place I would rather be than here.\",\n","  'speaker': 'PERSON',\n","  'context': [\"It's nice that all the people who are lonely on Valentine's Day can come here tonight and be together.\"],\n","  'context_speakers': ['RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I'm just inferring this is a couch because the evidence suggests the coffee table is having a tiny garage sale.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"Great Caesar's ghost, look at this place.\",\n","   \"So Penny's a little messy.\",\n","   'A little messy?',\n","   'The Mandelbrot set of complex numbers is a little messy. This is chaos.',\n","   'Excuse me.',\n","   'Explain to me an organizational system where a tray of flatware on a couch is valid.'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'No. The dark crescent-shaped patterns under your arms conceal it nicely.',\n","  'speaker': 'SHELDON',\n","  'context': ['How do I look?',\n","   'Could you be more specific?',\n","   \"Can you tell I'm perspiring a little?\"],\n","  'context_speakers': ['LEONARD', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Yeah, my parents felt that naming me Leonard and putting me in Advanced Placement classes wasn't getting me beaten up enough.\",\n","  'speaker': 'LEONARD',\n","  'context': ['He switched over to high-energy radiation research, had a little mishap, and now the other guys are uncomfortable sitting next to him.',\n","   \"You're in?\",\n","   'Yeah, sure, why not?',\n","   'Great, we rehearse on Tuesdays at your place.',\n","   'Why at my place?',\n","   'Department of Energy said our regular space is kind of a hot zone.',\n","   'Nice meeting you.',\n","   \"Yeah, you, too.\\nLeonard, I didn't know you played the cello.\"],\n","  'context_speakers': ['PERSON',\n","   'PERSON',\n","   'LEONARD',\n","   'PERSON',\n","   'LEONARD',\n","   'PERSON',\n","   'PERSON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Well, I'm sorry, too, but there's just no room for you in my wallet.\",\n","  'speaker': 'SHELDON',\n","  'context': ['This is Dr. Sheldon Cooper.',\n","   'Yeah, I need to cancel my membership to the planetarium.'],\n","  'context_speakers': ['SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Well, there's always the possibility that a trash can spontaneously formed around the letter but Occam's rasor would suggest that someone threw it out.\",\n","  'speaker': 'SHELDON',\n","  'context': ['I promised my mother.',\n","   'I suppose you could have your android penis circumcised.',\n","   \"But that's something your rabbi would have to discuss with the manufacturer.\",\n","   \"Not to mention you'd have to power down on saturdays.\",\n","   'Sheldon, why is this letter in the trash?'],\n","  'context_speakers': ['HOWARD', 'RAJ', 'RAJ', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Just the latest copy of Applied Particle Physics quarterly.',\n","  'speaker': 'SHELDON',\n","  'context': ['Hey Sheldon.', 'Hello Penny.', 'Get anything good?'],\n","  'context_speakers': ['PENNY', 'SHELDON', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'On a dusty highway just outside Bakersfield. Where do you think? On Mars!',\n","  'speaker': 'HOWARD',\n","  'context': ['Will you at least stipulate that Star Trek IV: The Voyage Home is inarguably the best?',\n","   'I have three words for you: Wrath of Khan.',\n","   \"Thank God, you're here.\",\n","   \"What's the emergency?\",\n","   'I got the Mars Rover stuck in a ditch.',\n","   'Where?'],\n","  'context_speakers': ['SHELDON',\n","   'RAJ',\n","   'HOWARD',\n","   'LEONARD',\n","   'HOWARD',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"You were right. Your larynx is terribly inflamed. I've never seen anything like it.\",\n","  'speaker': 'PERSON',\n","  'context': ['Oh no', 'What?'],\n","  'context_speakers': ['PERSON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Assuming I can keep up this pace, three hours, 11 minutes. Plus however long it takes to conclude this fairly pointless conversation.',\n","  'speaker': 'SHELDON',\n","  'context': ['See, by attaching RFID tags to my clothing, it will enable my laptop to read and identify the items with this wand.',\n","   'I will then cross-reference them against destination, anticipated activity spectrum, weather conditions, duration of trip, et cetera.',\n","   'That does sound much simpler.',\n","   'How long is this going to take?'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'LEONARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Obviously, waitressing at The Cheese Cake Factory is a complex socio-economic activity that requires a great deal of analysis and planning.',\n","  'speaker': 'SHELDON',\n","  'context': [\"My God, she didn't!\",\n","   'What could she possibly be talking about for so long?'],\n","  'context_speakers': ['PENNY', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'When I have enough to destroy all the human toast on the Battlestar known as Galactica.',\n","  'speaker': 'SHELDON',\n","  'context': ['When are you going to stop making Cylon toast?'],\n","  'context_speakers': ['LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"No, I mean Princess Leia. Of course I mean Bernadette. She's a wonderful girl and she really likes you.\",\n","  'speaker': 'PERSON',\n","  'context': ['Thank you.',\n","   'So, shall we get started?',\n","   'Sure.\\nBut can I ask you a question first?',\n","   'Do you want to play Cylon and colonist?',\n","   \"I want to know why you're playing make-believe with me when you could be out with a real woman tonight.\",\n","   'You mean, Bernadette?'],\n","  'context_speakers': ['HOWARD',\n","   'HOWARD',\n","   'PERSON',\n","   'HOWARD',\n","   'PERSON',\n","   'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'I understand your point, but given a choice, Jews always go with Chinese food.',\n","  'speaker': 'HOWARD',\n","  'context': [\"There's a 39.95 lover's special at P.F. Chang's.\",\n","   'Egg rolls, dumplings, bottomless wok, and you get your picture taken on the big marble horse out front.',\n","   \"Given that Saint Valentine was a third century Roman priest who was stoned and beheaded, wouldn't a more appropriate celebration of the evening be taking one's steady gal to witness a brutal murder?\"],\n","  'context_speakers': ['HOWARD', 'HOWARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Well, I always figured it was to study us, discover our weaknesses, and report back to your alien overlords.',\n","  'speaker': 'PENNY',\n","  'context': ['Hello.',\n","   'Hello',\n","   'Alright dispense with the friendlt banter',\n","   \"I believe you know why I'm here.\"],\n","  'context_speakers': ['SHELDON', 'PENNY', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, I'm so glad you asked it like that. You.\",\n","  'speaker': 'LEONARD',\n","  'context': [\"Hey, Sheldon. I was up in the administration office, and I happened to overhear the name of the winner of this year's Chancellor's Award for Science.\",\n","   'And you want to rub my nose in the fact that my contributions are being overlooked again?',\n","   'I am the William Shatner of theoretical physics.',\n","   \"All right, I'll play. What self-important, preening fraud are they honoring this year?\"],\n","  'context_speakers': ['LEONARD', 'SHELDON', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'In the bedrooms of ten-year- old girls, where they belong.',\n","  'speaker': 'SHELDON',\n","  'context': ['Wow, this place is awesome.',\n","   'Where do they keep the Archies?'],\n","  'context_speakers': ['PERSON', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Yes, I'd like a seven-day course of penicillin, some, uh, syrup of ipecac-- to induce vomiting-- and a mint.\",\n","  'speaker': 'SHELDON',\n","  'context': ['The backwash into this glass is every pathogen that calls your mouth home, sweet home.',\n","   'Not to mention the visitors who arrive on the dancing tongue of your subtropical girlfriend.',\n","   \"Hey! That's my sister and my country you're talking about.\",\n","   \"Leonard may have defiled one, but I won't have you talking smack about the other.\",\n","   'You guys ready to order?'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'RAJ', 'RAJ', 'BERNADETTE'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I'd like to respond to that sarcastically: Yes, I relish the thought of a stranger covering my body with oil and rubbing it.\",\n","  'speaker': 'SHELDON',\n","  'context': ['the organism responsible for Michelob Lite.',\n","   'Is there something wrong with your neck?',\n","   \"It's a little stiff.\",\n","   'What a remarkably fragile structure to support such a valuable payload.',\n","   'Not unlike balancing a Faberge egg on a Pixie Stick.',\n","   'Have you considered massage?'],\n","  'context_speakers': ['AMY', 'AMY', 'SHELDON', 'SHELDON', 'SHELDON', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"No, I'm just allergic to people who get Nobel Prizes for no good reason.\",\n","  'speaker': 'SHELDON',\n","  'context': ['That lecture was a waste of time.',\n","   'I made more accurate diagrams of the expansion of the early universe on the nursery wall with the contents of my diaper.',\n","   'Are you getting sick?'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Well, if you don't mind looking like an orange traffic cone, great.\",\n","  'speaker': 'BERNADETTE',\n","  'context': ['So, Bernadette, how goes the hunt for bridesmaid dresses?'],\n","  'context_speakers': ['PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'said horrible things about each other in public... all in all, a pretty magical night.',\n","  'speaker': 'LEONARD',\n","  'context': ['Okay. So... we went out, saw a movie, met some nice people,'],\n","  'context_speakers': ['LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'How\\'s this? \"Pleased to meet you, Dr. Gablehauser. \"How fortunate for you that the university\\'s \"chosen to hire you, despite the fact \"that you\\'ve done no original research in 25 years, \"and instead have written a series of popular books \"that reduce the great concepts of science \"to a series of anecdotes, \"each one dumbed down to accommodate the duration \"of an average bowel movement.',\n","  'speaker': 'SHELDON',\n","  'context': ['You know, I am not going to enjoy this party.',\n","   \"I know, I'm familiar with you.\",\n","   'The last department party, Professor Finkleday cornered me and talked about spelunking for 45 minutes.',\n","   'Yes, I was there.',\n","   \"You know what's interesting about caves?\",\n","   'What?',\n","   'Nothing.',\n","   \"Then we'll avoid him, we'll meet the new department head, congratulate him, shake his hand and go.\"],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Sadly, Mrs. Giselle Wolowitz is sensitive to chlorine. Lucky for her I like my fashion models pruny and bug-eyed.',\n","  'speaker': 'HOWARD',\n","  'context': ['Did I miss anything? Did they kick Giselle off?',\n","   'Not yet, but her underwater photo shoot was an embarrassment.'],\n","  'context_speakers': ['LEONARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"If you don't like this stuff, let's go next door and build her a bear.\",\n","  'speaker': 'RAJ',\n","  'context': [\"You're kidding. You've got lotions and bath oils and soaps. That's the estrogen hat trick.\",\n","   'What it is is a cacophonous assault of eucalyptus, bayberry, cinnamon and vanilla.',\n","   \"It's as if my head were trapped in the pajamas of a sultan.\"],\n","  'context_speakers': ['HOWARD', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Kind of makes you feel more alive, doesn't it?\",\n","  'speaker': 'PERSON',\n","  'context': ['but wait to be captured or killed.',\n","   \"That's the worst part, the waiting.\",\n","   \"All the while knowing that there's a paint pellet out there with your name on it.\",\n","   'The big wet ball of death.'],\n","  'context_speakers': ['HOWARD', 'PERSON', 'HOWARD', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'So? Do cocaine smugglers write \"cocaine\" on the box?',\n","  'speaker': 'SHELDON',\n","  'context': [\"They're here.\",\n","   \"Stay calm. We don't know anything about them yet.\",\n","   'What are you doing?',\n","   \"I'm checking for musical instruments.\",\n","   'Does that sound like castanets to you?',\n","   'The box says \"kitchen.\"'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"But if she dumps you, she'll have a new boyfriend by tomorrow morning and you'll have a new girlfriend when you figure out how to build one.\",\n","  'speaker': 'HOWARD',\n","  'context': ['Let me ask you something.',\n","   \"Do you think it's okay for Penny to have an ex-boyfriend sleep on her couch?\",\n","   \"No, I mean, she's obviously way out of line.\",\n","   'Thank you!'],\n","  'context_speakers': ['LEONARD', 'LEONARD', 'HOWARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Leonard's work is nearly as amazing as third graders growing lima beans in wet paper towels.\",\n","  'speaker': 'SHELDON',\n","  'context': ['I am.\\nAre you interested in physics?',\n","   \"Oh, I find it fascinating. If I hadn't gone into microbiology, I probably would have gone into physics.\\nOr ice dancing.\",\n","   \"Actually, my tests of the Aharonov-Bohm quantum interference effect have reached an interesting point. Right now, we're testing the phase shift due to an electric potential.\",\n","   \"That's amazing.\",\n","   'Yes.'],\n","  'context_speakers': ['LEONARD',\n","   'BERNADETTE',\n","   'LEONARD',\n","   'BERNADETTE',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Okay, Amy, you're being silly. I'm not concerned about who hangs out with who. And I certainly don't have a problem with Leonard's new girlfriend who wears way too much makeup. Ooh, these are cute. Of course if I buy them, I'll have to rent my womb out to a gay couple.\",\n","  'speaker': 'PENNY',\n","  'context': ['That bitch is crafty.',\n","   'You think?',\n","   'Of course. How does the cheetah attack a herd of wildebeests? By going after its weakest member.',\n","   'Well, what makes me the weakest member?',\n","   \"Your trusting nature coupled with your teeny-tiny body. You wouldn't last a minute on the Serengeti.\"],\n","  'context_speakers': ['AMY', 'BERNADETTE', 'AMY', 'BERNADETTE', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Right, right, whereas Jewish mothers take a casual, la-di-da approach to their sons.',\n","  'speaker': 'HOWARD',\n","  'context': ['I was thinking of weaving it into her eulogy.',\n","   \"Howard, you've got to tell your mother.\",\n","   \"Hey, have you told your parents you're dating this short glass of skim milk here?\",\n","   \"Uh... that's different.\\nFirst of all, we're not engaged; and second, Indian parents are very protective of their children.\"],\n","  'context_speakers': ['HOWARD', 'PERSON', 'HOWARD', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Your girlfriend and your ex-girlfriend are hanging out together? Oh, yeah, that can only be good for you.',\n","  'speaker': 'RAJ',\n","  'context': ['You mean when he was little.',\n","   \"Not as little as you'd want him to be.\",\n","   'You know, this stuff is horrible.\\nYou want to see if we can find the cafeteria and get real coffee? Sure.',\n","   \"We're going to the cafeteria to get some coffee. \\nYou want anything?\",\n","   \"I'm fine.\",\n","   \"That's nice that they're getting along.\"],\n","  'context_speakers': ['PENNY', 'PERSON', 'PENNY', 'PERSON', 'LEONARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'You see betrayal in others, but not yourself.',\n","  'speaker': 'SHELDON',\n","  'context': ['What choice did I have? She went behind my back and turned my own mother against me.',\n","   \"Wow. You're not only our first astronaut.\\nYou're also the first one of us to kick a girl out of bed.\",\n","   \"You're like a rock star.\",\n","   'Little bit.',\n","   'I hate to say it, but she did kind of betray you.'],\n","  'context_speakers': ['HOWARD', 'RAJ', 'RAJ', 'HOWARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'You know, the holidays are just around the corner. Maybe he wants to see if he can lodge the other testicle up there.',\n","  'speaker': 'SHELDON',\n","  'context': ['No, that was a different, different guy.',\n","   'Was he the one who used your head to open a nut?',\n","   'Oh, oh, oh.',\n","   'Was he the one who made you eat your arm hair?',\n","   \"No, but, actually, that was this guy's sister.\",\n","   'All right, well, what do you think he wants?',\n","   \"I don't know.\"],\n","  'context_speakers': ['LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'LEONARD',\n","   'PENNY',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Yeah, terrific. The other astronauts would love to go hurtling through space with a guy named \"Crash.\"',\n","  'speaker': 'HOWARD',\n","  'context': ['Buzz Lightyear is not real.',\n","   \"No, that's not what I'm talking about.\",\n","   'Well, are you talking about when he thought he was real?',\n","   'No',\n","   'Okay, um... how about \"Crash\"? \\nHoward \"Crash\" Wolowitz.'],\n","  'context_speakers': ['RAJ', 'HOWARD', 'RAJ', 'HOWARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Also instead of just living in your mother's house, you could actually live inside her body.\",\n","  'speaker': 'LEONARD',\n","  'context': ['Wrong. Leonard?',\n","   'Horse, but mostly just for the height.',\n","   'A little bit for the genital girth.',\n","   \"Wrong, and let's keep it clean, shall we?\",\n","   \"Kangaroo, uh, I'd be Kanga-Jew.\",\n","   'The first of my people to dunk a basketball.'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'LEONARD',\n","   'SHELDON',\n","   'HOWARD',\n","   'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Well, if that's all it takes, it's a good thing you don't have a dog.\",\n","  'speaker': 'AMY',\n","  'context': ['You slept with him?',\n","   \"I didn't know what else to do.\\nHe had those big, sad eyes.\",\n","   'Oh, sure, you had no choice.',\n","   'He looked at me like this.'],\n","  'context_speakers': ['AMY', 'PENNY', 'BERNADETTE', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Like Burt and Ernie. You guys even teach me stuff about words and numbers.',\n","  'speaker': 'PENNY',\n","  'context': ['I swear, that man is the most egotistical, insufferable human being I have ever met.',\n","   'Yeah, but you two make such a cute couple.'],\n","  'context_speakers': ['LEONARD', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'I enjoyed it. When you told me I was going to be \"losing my virginity,\" I didn\\'t think you meant showing me <i>Raiders of the Lost Ark</i> for the first time.',\n","  'speaker': 'AMY',\n","  'context': [\"So what'd you think?\", 'It was good.', \"That's it?\", '\"Good\"?'],\n","  'context_speakers': ['SHELDON', 'AMY', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'It was very entertaining... despite the glaring story problem.',\n","  'speaker': 'AMY',\n","  'context': ['I enjoyed it. When you told me I was going to be \"losing my virginity,\" I didn\\'t think you meant showing me Raiders of the Lost Ark  for the first time.',\n","   'My apologies. I chose my words poorly.\\nI should have said you were about to have your world rocked on my couch.',\n","   \"Anyway, thank you for watching it.\\nIt's one of my all-time favorites.\"],\n","  'context_speakers': ['AMY', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Sure, cut the foreigner in half. There's a billion more where he came from.\",\n","  'speaker': 'RAJ',\n","  'context': [\"And I cannot believe people pay for horoscopes, but on a more serious note, it's 8:13 and we're still not playing Halo.\",\n","   \"Fine. We'll just play one-on-one until he gets back.\",\n","   \"One-on-one? We don't play one-on-one.\",\n","   'We play teams, not one-on-one.',\n","   'One-on-one.',\n","   'The only way we can play teams at this point is if we cut Raj in half.'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Oh, please. The only way she could make a contribution to science would be if they resumed sending chimps into space.',\n","  'speaker': 'SHELDON',\n","  'context': ['Pastries such as bear claws?',\n","   'Yeah, sure.',\n","   \"I don't like bear claws.\",\n","   'Hey, Penny.',\n","   'Dumbass.',\n","   'Leslie Winkle.',\n","   'Of all the overrated physicists in all the labs in all the world, why does it have to be Leslie Winkle?',\n","   \"They have a lot in common. I mean, they're both scientists.\"],\n","  'context_speakers': ['SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'PERSON',\n","   'PERSON',\n","   'SHELDON',\n","   'SHELDON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Ooh, big talk from a man who was once treed by a chicken.',\n","  'speaker': 'RAJ',\n","  'context': [\"Don't push me, Sheldon.\\nI may be small, but I took kung fu when I was 13, and I remember a good deal of it.\",\n","   'Oh, really?',\n","   'Well, I grew up with an older brother and a very contentious twin sister.',\n","   'And I believe I could easily best you in any physical confrontation-- be it noogies, swirlies or the classic \"Why are you hitting yourself?\"'],\n","  'context_speakers': ['HOWARD', 'SHELDON', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, sure. And while we're at it, why don't we put our hands behind our backs, have an old-fashioned eating contest?\",\n","  'speaker': 'SHELDON',\n","  'context': ['I told my girlfriend Bernadette she could join us for dinner.',\n","   'Sure. The more, the merrier.',\n","   \"Wa- no, that's a false equivalency. More does not equal merry.\",\n","   \"If there were 2,000 people in this apartment right now, would we be celebrating?\\nNo, we'd be suffocating.\",\n","   'Sheldon...',\n","   'Don\\'t \"Sheldon\" me. We ordered for five people, not six.',\n","   \"Oh, come on, it's fine. We'll just put it all on the table, you know, family style.\"],\n","  'context_speakers': ['HOWARD',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, you know, that is so weird that yours came and mine didn't.\",\n","  'speaker': 'PENNY',\n","  'context': ['Hey Sheldon.',\n","   'Hello Penny.',\n","   'Get anything good?',\n","   'Just the latest copy of Applied Particle Physics quarterly.'],\n","  'context_speakers': ['PENNY', 'SHELDON', 'PENNY', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Okay, here's an idea. What if I change my name and go live with my cousin and her husband, Avi, in Israel?\",\n","  'speaker': 'HOWARD',\n","  'context': [\"It's actually a very elegant system. You know, what goes around comes around.\",\n","   'Speaking of what goes around comes around...',\n","   'Okay, look. Instead of trying to reinforce this structure here, what if we just ran another line, bypass it entirely?',\n","   \"It won't work. The diameter of the tubing is insufficient.\",\n","   'What if we reposition the collection tank?',\n","   \"It won't work. No way to mount it.\"],\n","  'context_speakers': ['RAJ', 'HOWARD', 'RAJ', 'SHELDON', 'RAJ', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Can we please talk about something else? Maybe something vaguely related to life as we know it on this planet?',\n","  'speaker': 'LEONARD',\n","  'context': [\"I honestly don't care.\",\n","   \"Really? Because every time we've talked about unusual animal genitals, you've always had some pretty strong and controversial opinions.\",\n","   \"What do you want from me? I just don't give a rat's ass.\",\n","   \"Would that be a giant rat's ass?\",\n","   'For the record, giant rats are possible.'],\n","  'context_speakers': ['LEONARD', 'RAJ', 'LEONARD', 'HOWARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'A lot of people are working on that research. So what is all this?',\n","  'speaker': 'LEONARD',\n","  'context': ['And this is a schematic for a bird-trapping device that will ultimately backfire and cause me physical injury?',\n","   \"What I'm doing here is trying to determine when I'm going to die.\"],\n","  'context_speakers': ['SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Oh, and the science dudes want to talk science with you. What do you want to talk about-- rocks, dinosaurs, our friend the beaver?',\n","  'speaker': 'HOWARD',\n","  'context': [\"That's what I love about science-- there's no one right answer.\",\n","   'So...you and Zack again, huh?',\n","   'Yeah, yeah, me and Zack again.',\n","   'Were we here earlier?',\n","   'Okay, we should, we should go.',\n","   'Ah, not yet. I want to talk science with the science dudes.'],\n","  'context_speakers': ['PERSON',\n","   'LEONARD',\n","   'PENNY',\n","   'PERSON',\n","   'PENNY',\n","   'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Hey, you taught me something. Who would have thought it?',\n","  'speaker': 'SHELDON',\n","  'context': [\"Why not? It seems like you're improvising your entire curriculum.\",\n","   'This is all about listening and responding.',\n","   'Gotcha.',\n","   \"I'm going to create a character and a situation, and you just jump in when you feel it.\",\n","   'All right.',\n","   'All right.',\n","   'Action.',\n","   'Okay, it\\'s not a movie. It\\'s improv. So no one calls \"action.\"'],\n","  'context_speakers': ['SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'An entire dinner to talk about your research? Where you going, the drive-thru at Jack in the Box?',\n","  'speaker': 'SHELDON',\n","  'context': [\"Yes, I live with him.\\nI don't... I-I really don't know why.\",\n","   \"Tonight? Sure, that'd be great. Okay, I'll see you then. Bye.\",\n","   'She wants to have dinner and talk about my research.'],\n","  'context_speakers': ['LEONARD', 'LEONARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Saying that while holding a snowflake is a bit heavy-handed, don't you think?\",\n","  'speaker': 'AMY',\n","  'context': ['Um, no.',\n","   'You should get one.',\n","   \"What's that?\",\n","   'It appears to be something preserved between two pieces of acrylic.',\n","   \"Oh, it's a snowflake from the North Pole. Leonard gave it to me last year.\",\n","   'Oh, God, that is so romantic.',\n","   'Yeah, it was.',\n","   \"Leonard's really one of a kind.\"],\n","  'context_speakers': ['BERNADETTE',\n","   'AMY',\n","   'BERNADETTE',\n","   'AMY',\n","   'PENNY',\n","   'BERNADETTE',\n","   'PENNY',\n","   'BERNADETTE'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'You mean for a person whose neck massages feel like an eagle is trying to carry you to its nest?',\n","  'speaker': 'LEONARD',\n","  'context': ['This is amazing.\\nHow did you even get this idea?',\n","   'I called your dad.\\nI asked him what things you liked to do when you were a kid.',\n","   'This seemed easier than getting a cow out here so you could tip it.',\n","   \"Okay, you're kind of really great.\"],\n","  'context_speakers': ['PENNY', 'LEONARD', 'LEONARD', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'Dumped her? What, did he use a forklift?',\n","  'speaker': 'RAJ',\n","  'context': ['to rebrand the stuff as a cure for constipation.',\n","   'Way to make lemonade.\\nYou know, from around the corner where fudge is made.',\n","   'Hey, I got a favor to ask.',\n","   'Sure.',\n","   \"My mom's been kind of an emotional wreck since that dentist she was dating dumped her.\"],\n","  'context_speakers': ['HOWARD', 'RAJ', 'HOWARD', 'RAJ', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'How nice. Maybe they can carpool when they visit you in federal prison.',\n","  'speaker': 'SHELDON',\n","  'context': ['So my friend Leonard is gonna take you home.',\n","   \"Okay. Let's go, friend Leonard.\",\n","   \"I'll call ya.\",\n","   'So are you a scientist like Howard?',\n","   \"No one's a scientist like Howard.\",\n","   'My mother is so gonna love her.'],\n","  'context_speakers': ['HOWARD',\n","   'PERSON',\n","   'HOWARD',\n","   'PERSON',\n","   'LEONARD',\n","   'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"When I didn't pay my bill, the Department of Water and Power thought I would enjoy the ambience.\",\n","  'speaker': 'PENNY',\n","  'context': ['Hey.', 'Hey.', 'A little mood lighting, huh?'],\n","  'context_speakers': ['LEONARD', 'PENNY', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': 'In 1917, when Albert Einstein established the theoretic foundation for the laser in his paper <i>\"Zur Quantentheorie der</i> <i>Strahlung,\"</i> his fondest hope was that the resultant device be bitchin\\'.',\n","  'speaker': 'SHELDON',\n","  'context': [\"Happy now? I'm moving the dirty sock.\",\n","   'Thank you. Raj, keep an eye out for the other one.',\n","   'Hey, guys, this is my friend Zack.',\n","   'Hey.',\n","   'Hello.',\n","   \"Whoa! Is that the laser? It's bitchin'.\",\n","   'Yes.'],\n","  'context_speakers': ['HOWARD',\n","   'SHELDON',\n","   'PENNY',\n","   'PERSON',\n","   'SHELDON',\n","   'PERSON',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I'm sorry, fiddling with yourself in the bathtub is a real class act.\",\n","  'speaker': 'PERSON',\n","  'context': ['So nice you could join me this evening.',\n","   \"You're looking lovely as always.\",\n","   'Thanks, Howard.  Always nice to be part of your masturbatory fantasies.',\n","   \"Come on, Katee. Don't make it sound so cheap.\"],\n","  'context_speakers': ['HOWARD', 'HOWARD', 'PERSON', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"I think I'm gonna go. Thank you for the burrito and the pork rinds and the 20-minute lecture on why monster trucks are better than regular trucks.\",\n","  'speaker': 'PERSON',\n","  'context': [\"Probably because I didn't read the instructions.\\nNo wrapper's going to tell me what to do.\",\n","   'Unless his name is Jay-Z.',\n","   'Is something going on?',\n","   'What do you mean?',\n","   \"Well, you're acting all weird.\\nAnd I'm pretty weird, so I think I know what I'm talking about.\",\n","   \"I don't know. Maybe it's all those steroids I've been taking.\"],\n","  'context_speakers': ['RAJ', 'RAJ', 'PERSON', 'RAJ', 'PERSON', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': True},\n"," {'utterance': \"Since it's not bee season, you can have my epinephrine.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Here we go. Pad thai, no peanuts.',\n","   'But does it have peanut oil?',\n","   \"I'm not sure. Everyone keep an eye on Howard in case he starts to swell up.\"],\n","  'context_speakers': ['LEONARD', 'HOWARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Leonard, may I present, live from New Delhi, Dr. and Mrs. V. M. Koothrappali.',\n","  'speaker': 'RAJ',\n","  'context': ['Do you really need the Honorary Justice League of American membership card?',\n","   \"It's been in every wallet I've owned since I was five.\",\n","   'Why?',\n","   'It says, \"Keep this on your person at all times.\"',\n","   \"It's right here under Batman's signature.\",\n","   \"and this is Leonard and Sheldon's apartment.\",\n","   'Guess whose parents just got broadband.'],\n","  'context_speakers': ['LEONARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'RAJ',\n","   'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I've got some feelers out. In the meantime, listen to this.\",\n","  'speaker': 'RAJ',\n","  'context': [\"but it was between you and the Museum of Natural History, and frankly, you don't have dinosaurs.\",\n","   \"I'll miss you, too. Bye-bye.\",\n","   \"Okay, I know you're texting about me and I'd really like you to stop.\",\n","   'Oh, dear, I am rightly and truly screwed.',\n","   'Hey, I thought you were finding new friends.'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'RAJ', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Answering the question once and for all why Wolowitz bailed out of medical school.',\n","  'speaker': 'RAJ',\n","  'context': ['Can you open this for me?',\n","   'Can I see your stitches?',\n","   'Sure.'],\n","  'context_speakers': ['LEONARD', 'RAJ', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Terrific, it's just that... I don't think Penny wants me to go.\",\n","  'speaker': 'LEONARD',\n","  'context': [\"I don't think I can go to the North Pole.\",\n","   \"I know you're concerned about disappointing me but I want you to take comfort from the knowledge that my expectations of you are very low.\",\n","   \"Yeah, that's very comforting.\",\n","   \"Comforting is a part of leadership.\\nIt's not a part I care for, but such is my burden.\"],\n","  'context_speakers': ['LEONARD', 'SHELDON', 'LEONARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"- Chocolate? - No, I don't want any chocolate!\",\n","  'speaker': 'LEONARD',\n","  'context': [\"You'll never guess who they got to replace you at work.\",\n","   \"Okay, I know what you're doing.\",\n","   'Really?',\n","   \"Yes, you're using chocolates as positive reinforcement for what you consider correct behaviour.\\nVery Good.\"],\n","  'context_speakers': ['LEONARD', 'SHELDON', 'LEONARD', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Okay, I got a bone to pick with you.',\n","  'speaker': 'HOWARD',\n","  'context': [\"That's why Leonard does it.\",\n","   'Okay, I just have one question. What exactly are sub-atomic particles?',\n","   'A good question.',\n","   'Thank you.',\n","   'And to answer it, we first must ask ourselves: \"What is physics?\" Oh, balls. It\\'s a warm summer evening in ancient Greece.'],\n","  'context_speakers': ['SHELDON', 'PENNY', 'SHELDON', 'PENNY', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Excuse me, I'm thirsty, so I'm going to go to the refrigerator and get myself a refreshing beverage.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"Oh, there it is: now that you don't have a girlfriend, you want to hang out with me again.\",\n","   \"I never stopped hanging out with you. Oh, please, we all know I'm the friend you call when you have no other options. If we were the Justice League, I'd be Aquaman.\",\n","   'I wish you were Aquaman. Then I could send you to scoop my mom out of the old lady tank.'],\n","  'context_speakers': ['RAJ', 'LEONARD', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'I just want to ask you a few questions about Howard Wolowitz.',\n","  'speaker': 'PERSON',\n","  'context': [\"Yes.\\nI'm Special Agent Page, FBI.\",\n","   \"You say you're Special Agent Page, FBI.\",\n","   \"Here's my I.D.\",\n","   \"And here is my Justice League membership card. But that doesn't prove I know Batman.\"],\n","  'context_speakers': ['PERSON', 'SHELDON', 'PERSON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Yes. In fact, our menses are synchronized. Penny. Bernadette tells me you're planning a girls' night.\",\n","  'speaker': 'AMY',\n","  'context': [\"Besides, Penny and I are having a girls' night tonight.\",\n","   \"Girls' night? What does that entail?\",\n","   'Oh, you know, girls get together, hang out, share girl talk.',\n","   \"I'm a girl.\",\n","   \"Oh. Well, maybe you can join us. I'll ask Penny.\",\n","   'No need. Penny and I are very close.',\n","   'You are?'],\n","  'context_speakers': ['BERNADETTE',\n","   'AMY',\n","   'BERNADETTE',\n","   'AMY',\n","   'BERNADETTE',\n","   'AMY',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': '...so anyway, to make a long story short, turns out I have an unusually firm cervix.',\n","  'speaker': 'AMY',\n","  'context': [\"And here's where we are. The runts in a large litter.\\nunlikely to ever reach the nourishing teats of Indiana Jones.\",\n","   \"So I guess it's a good thing we stopped for dinner.\",\n","   'You know, guys, when facing disappointment, Eat, Pray, Love  teaches us',\n","   'Oh, shut up.'],\n","  'context_speakers': ['SHELDON', 'LEONARD', 'RAJ', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Yeah, but you were just sleeping, because I forboded you to have sex.',\n","  'speaker': 'RAJ',\n","  'context': [\"That's very nice of you, but I made other plans.\",\n","   'Well, just keep in mind that should you ever need a slightly apathetic tertiary friend, I stand at the ready.',\n","   'Hi, Howard. Hi, Sheldon.',\n","   'Hello. Hey.',\n","   'Why are you holding hands? I forbid you to hold hands.',\n","   'Rajesh, you know Leonard and I spent the night together.'],\n","  'context_speakers': ['HOWARD',\n","   'SHELDON',\n","   'PERSON',\n","   'SHELDON',\n","   'RAJ',\n","   'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'I gargled with tequila and may have swallowed a teeny bit.',\n","  'speaker': 'SHELDON',\n","  'context': [\"Hey, have you told your parents you're dating this short glass of skim milk here?\",\n","   \"Uh... that's different.\\nFirst of all, we're not engaged; and second, Indian parents are very protective of their children.\",\n","   'Right, right, whereas Jewish mothers take a casual, la-di-da approach to their sons.'],\n","  'context_speakers': ['HOWARD', 'PERSON', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Here's some other fun news on the Raj/Emily front. He gave her a pair of diamond earrings and leased her a car.\",\n","  'speaker': 'HOWARD',\n","  'context': ['because you no longer wear underpants.',\n","   'The dice giveth and the dice taketh away.',\n","   'Is Raj out with Emily again?',\n","   'Yeah, every night for the last month.',\n","   \"Wow, can't believe he has a girlfriend.\",\n","   'Me neither.'],\n","  'context_speakers': ['LEONARD',\n","   'SHELDON',\n","   'PENNY',\n","   'LEONARD',\n","   'PENNY',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I don't know what I'm going to do. I don't want to break up over this.\",\n","  'speaker': 'BERNADETTE',\n","  'context': [\"He can't go to space.\",\n","   \"He's like a baby bird.\",\n","   'Do you know he once got an asthma attack from reading an old library book?',\n","   \"You're kidding.\",\n","   'No, I was there that day.',\n","   'Sheldon threw his back out handing him that book.'],\n","  'context_speakers': ['BERNADETTE',\n","   'BERNADETTE',\n","   'BERNADETTE',\n","   'AMY',\n","   'PENNY',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"We don't say that, either. I'll make you a list. Oh, that would be mighty white of you.\",\n","  'speaker': 'LEONARD',\n","  'context': ['Three, sadly, are dead.',\n","   'Mom, you remember Rajesh?',\n","   'Rajesh, my mother. Of course.',\n","   'Mrs. Cooper. So nice to see you again.',\n","   \"Well, it's so nice to see you, too.\",\n","   'I thought it was our Indians that had the occasional alcohol problem.'],\n","  'context_speakers': ['RAJ', 'SHELDON', 'SHELDON', 'RAJ', 'PERSON', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'You have a thing of paper clips right there.',\n","  'speaker': 'RAJ',\n","  'context': [\"It is, isn't it?\",\n","   '(gasps) Oh, bother.',\n","   \"Isn't that just always the way? \\nYou go to staple something, and you're out of staples.\",\n","   \"Gosh, I wish I'd known that earlier today when I was at Staples.\"],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Well, that wady took high-wes pictures of my junk wast night for Cwaigswist.',\n","  'speaker': 'PERSON1',\n","  'context': ['(chimes) Wisten to me.\\nNot westauwant; westauwant.',\n","   '(chimes) I don\\'t know what you mean by \"not westauwant; westauwant.\"',\n","   'See? Total cwap.\\nYou suck, Siwi.',\n","   \"Hey, don't talk to her like that. \\nShe's a lady.\"],\n","  'context_speakers': ['PERSON1', 'PERSON2', 'PERSON1', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Maybe that's because I'm not being challenged. It's the same reason Einstein failed math.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"And I'm super smart, so it's probably true.\",\n","   \"Hey, I've been training in the field of neurobiology for 12 years.\",\n","   \"You've been here for three hours, and you've spent one of them in the bathroom.\",\n","   \"I'm sorry! It takes me a while to get things going on an unfamiliar toilet.\",\n","   \"Sheldon, I've given you the simplest things to do, and you haven't done one of them right.\"],\n","  'context_speakers': ['SHELDON', 'AMY', 'AMY', 'SHELDON', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"...to all the good times we had, like, uh, when we went camping and spent that night telling each other all our secrets. I told him I'm addicted to pedicures and he told me he lost his virginity to his cousin.\",\n","  'speaker': 'RAJ',\n","  'context': ['This man became my whole world.',\n","   'Yeah, nice speech, Francine.',\n","   \"I'm not done, but thank you.\",\n","   'I think back....'],\n","  'context_speakers': ['RAJ', 'PERSON', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Not exactly. Listen, if you don't mind, I'm not really up for chatting. I'm just going to sit here quietly and let my life flash before my eyes.\",\n","  'speaker': 'HOWARD',\n","  'context': ['and she makes me Froot Loops.',\n","   'Go with gay story, people are more accepting of that.',\n","   'Actually, Froot Loops just got married to a girl.',\n","   'Congratulations.',\n","   'Thanks, we decided to do it before the launch.',\n","   'You and Mrs. Loops have a big wedding?'],\n","  'context_speakers': ['HOWARD',\n","   'PERSON1',\n","   'PERSON2',\n","   'PERSON1',\n","   'HOWARD',\n","   'PERSON1'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Okay, the NASA Web site says Howard's final descent has begun.\",\n","  'speaker': 'RAJ',\n","  'context': [\"or Leonard's gym membership.\",\n","   \"We're not kicking him out. \\nStuart and I have become good friends.\",\n","   \"Okay, one vote for, one vote against.\\nLeonard, you're the tiebreaker.\",\n","   \"I don't have a problem with Stuart.\\nBesides, he gives us a 20% discount at his comic book store.\",\n","   \"Well, I don't sell my friendship that cheaply.\",\n","   'I can go 30.',\n","   'Welcome aboard, old chum.'],\n","  'context_speakers': ['SHELDON',\n","   'RAJ',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'PERSON',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Yeah, well, Bernadette's a little under the weather and my mom's kind of under my dentist.\",\n","  'speaker': 'HOWARD',\n","  'context': ['Oh, look at you.\\nYou, you look like you grew.',\n","   \"Yeah.\\nThe lack of gravity did decompress my spine, so I'm like an inch and a half taller.\",\n","   \"I'm going to the DMV tomorrow to get my license changed before I shrink back.\",\n","   'Oh, uh, come in, come in.',\n","   \"Oh, I, uh, I didn't think I was going to get to see you until tomorrow.\"],\n","  'context_speakers': ['RAJ', 'HOWARD', 'HOWARD', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"It's okay, you know, we space cowboys don't do what we do for glory and fame. We leave that to your rock stars and your athletes and your Howie Mandels.\",\n","  'speaker': 'HOWARD',\n","  'context': [\"Yeah, well, Bernadette's a little under the weather and my mom's kind of under my dentist.\",\n","   'Wait, your, your mother is sleeping with your dentist?',\n","   'Former dentist.\\nI need a new one now that I know where his hands have been.',\n","   \"So you're wandering all around by yourself?\\nThat's not the kind of hero's welcome an astronaut should come home to.\"],\n","  'context_speakers': ['HOWARD', 'RAJ', 'HOWARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'that sounds a little funny to an American ear.',\n","  'speaker': 'PERSON',\n","  'context': ['Wow, so you guys are like buds now?',\n","   'Oh, yeah, we hang out all the time.',\n","   \"Plus, he doesn't have a girlfriend, I don't have a girlfriend.\",\n","   \"It's like we both had these holes in our lives, but now we fill each other's holes.\"],\n","  'context_speakers': ['HOWARD', 'RAJ', 'PERSON', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': '\"To Stuart, your comic book store is out of this world. Just like the guy in this picture was.\"',\n","  'speaker': 'PERSON',\n","  'context': ['Oh, hey, Stuart, I got you a little souvenir from my trip to space.',\n","   \"Well, Howard, that's very nice of you.\",\n","   'Yeah, maybe. Open it first.',\n","   \"It's my official NASA portrait.\"],\n","  'context_speakers': ['HOWARD', 'PERSON', 'LEONARD', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Amy, I can't just ask Wil Wheaton to leave. He's a minor celebrity. Once you explain who he is, many people recognize him.\",\n","  'speaker': 'SHELDON',\n","  'context': ['What do you think?',\n","   'Can I speak to you for a second?',\n","   \"I'll be right back.\",\n","   'Feel free to play with yourself.',\n","   \"I don't care for your friend, he's being rude to me.\",\n","   'You need to ask him to leave.'],\n","  'context_speakers': ['SHELDON', 'AMY', 'SHELDON', 'SHELDON', 'AMY', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Anyway, to this day, I still can't see a box of crayons without crossing my legs.\",\n","  'speaker': 'PENNY',\n","  'context': ['Damn.',\n","   \"Well, at least when we do make love, Sheldon won't be thinking about his mother.\",\n","   \"And yes, that is a cleverly veiled reference to Howard's lifelong obsession to crawl back into her spacious womb.\"],\n","  'context_speakers': ['PENNY', 'AMY', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"And I'm cool with surprises. But nothing on a Jumbotron. I don't want to cry on a big screen like that.\",\n","  'speaker': 'LEONARD',\n","  'context': ['But if someday you decide you want to get married, you have to propose to me.',\n","   'Really?',\n","   \"Yes. It's all on you.\",\n","   'But I got to tell you, when the time comes, I want the whole nine yards.',\n","   'I want you down on one knee, flowers.\\nI want to be swept off my feet.',\n","   'Yeah, you got it.'],\n","  'context_speakers': ['LEONARD',\n","   'PENNY',\n","   'LEONARD',\n","   'LEONARD',\n","   'LEONARD',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"We could go to Red Lobster and talk about it. We're all thinking it. I just had the decency to wait for him to leave.\",\n","  'speaker': 'BERNADETTE',\n","  'context': ['The surf and turf sounds good.\\nSee if they corn on the cob.',\n","   \"Guys!\\nWe're trying to get him out of here, not you in.\",\n","   \"You're wasting your time.\\nJust please, leave me alone.\",\n","   'Now what?',\n","   \"I don't know.\"],\n","  'context_speakers': ['LEONARD', 'BERNADETTE', 'RAJ', 'HOWARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Oh, I think we were looking at creepy in the rearview mirror when I put up that camera.',\n","  'speaker': 'HOWARD',\n","  'context': ['Roger is Raj.',\n","   'I always thought, if I had a white name, it would be Gavin.',\n","   'Keep reading. What does it say?',\n","   \"No, no, I don't know if I should. \\nIf she wanted me to know about her blog, she would've told me.\",\n","   \"It's almost like I'm reading her diary.\",\n","   \"That's exactly what it's like. Keep reading.\",\n","   'No, this is creepy.'],\n","  'context_speakers': ['HOWARD',\n","   'RAJ',\n","   'HOWARD',\n","   'RAJ',\n","   'RAJ',\n","   'HOWARD',\n","   'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'A lane frequented by liars. Like you, you big liar!',\n","  'speaker': 'RAJ',\n","  'context': ['I just Googled \"foo-foo little dogs.\"',\n","   \"(Skype ringing) It's Raj. Stay quiet.\",\n","   '(chuckles): Hey!\\nBad timing.\\nBernadette just took Cinnamon out for a walk.',\n","   \"Hmm. Interesting. \\nDid they take a walk down Liars' Lane?\",\n","   'What?'],\n","  'context_speakers': ['HOWARD', 'HOWARD', 'HOWARD', 'RAJ', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I'm not going to bother him, I'm going to talk to him.\",\n","  'speaker': 'SHELDON',\n","  'context': ['I guess they really do read their mail.',\n","   \"Hey, isn't that Professor Proton?\",\n","   'Oh, yeah.',\n","   \"Look at him, just standing in line like he wasn't moderately famous 30 years ago.\",\n","   \"- Let's go say hello.\",\n","   \"- Oh, maybe we shouldn't bother him.\"],\n","  'context_speakers': ['LEONARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Can we talk about something else? Do you know that I yodel?',\n","  'speaker': 'SHELDON',\n","  'context': ['Yes.',\n","   'And some people in the science community are calling it \"The Wonder Blunder.\"',\n","   \"Who? Give me their names.\\nI bet it's Wolowitz.\",\n","   \"It's just such a fascinating story. Your calculations are way off, but they find the element anyway.\",\n","   \"It's like misreading a treasure map and still finding the treasure.\"],\n","  'context_speakers': ['SHELDON', 'PERSON', 'SHELDON', 'PERSON', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I can't believe we wasted all that time on our Hulk costumes.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"You know, there-there's always WonderCon in Anaheim, you know?\",\n","   \"That-That's just as good.\",\n","   'Excuse me.',\n","   'Oh, guys, this is really sad.',\n","   'And in a different way than it was 20 minutes ago.'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'PENNY', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I'm with him three years, nothing. She's with two minutes, and he's taking his pants off.\",\n","  'speaker': 'AMY',\n","  'context': ['to be a pharmaceutical sales rep.',\n","   'Maybe she can practice on you some time.',\n","   'Oh, it would be great to practice on a real doctor.',\n","   \"Yeah. I'm sure that'd be fine.\",\n","   'I have some odd freckles on my buttocks.',\n","   'Can I make an appointment for you to look at them?',\n","   'Um... okay, I guess.'],\n","  'context_speakers': ['BERNADETTE',\n","   'BERNADETTE',\n","   'PENNY',\n","   'PERSON',\n","   'SHELDON',\n","   'SHELDON',\n","   'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I'll give you a plan right now. Step one: open comic book store. Step two: start rumor this comic book store gives you genital warts.\",\n","  'speaker': 'SHELDON',\n","  'context': ['For starters, they shed and bite.',\n","   \"Hey, here's a thought.\\nWhy don't we put up the rest of the money that Stuart needs?\",\n","   \"So, we'd be, like, owners of a comic book store?\",\n","   \"It's kind of a dream come true.\",\n","   'That does sound fun.',\n","   'Ooh, maybe we could come up with a business plan to compete with this place.'],\n","  'context_speakers': ['SHELDON',\n","   'LEONARD',\n","   'RAJ',\n","   'LEONARD',\n","   'HOWARD',\n","   'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Lois Lane is falling, accelerating at an initial rate of 32 feet per second per second. Superman swoops down to save her by reaching out two arms of steel. Miss Lane, who is now traveling at approximately 120 miles an hour, hits them and is immediately sliced into three equal pieces.',\n","  'speaker': 'SHELDON',\n","  'context': ['A marathon? How many Superman movies are there?',\n","   \"You're kidding, right?\",\n","   'I do like the one where Lois Lane falls from the helicopter and Superman swooshes down and catches her. Which one was that?',\n","   'One.',\n","   'You realize that scene was rife with scientific inaccuracy.',\n","   \"Yes, I know, men can't fly.\",\n","   \"Let's assume that they can.\"],\n","  'context_speakers': ['PENNY',\n","   'SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'SHELDON',\n","   'PENNY',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Sheldon, if you were a robot, and I knew and you didn't...\",\n","  'speaker': 'HOWARD',\n","  'context': ['On the other hand, no arousal.',\n","   'None?',\n","   'None.',\n","   'Thank you for your time.',\n","   'Thank you.',\n","   'None at all?'],\n","  'context_speakers': ['PERSON',\n","   'LEONARD',\n","   'PERSON',\n","   'LEONARD',\n","   'PERSON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Not to mention you'd have to power down on saturdays.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"I'm with you. I just have to make sure if I'm a synthetic human I'd still be Jewish.\",\n","   'I promised my mother.',\n","   'I suppose you could have your android penis circumcised.',\n","   \"But that's something your rabbi would have to discuss with the manufacturer.\"],\n","  'context_speakers': ['HOWARD', 'HOWARD', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'I think tonight was a very good start.',\n","  'speaker': 'LEONARD',\n","  'context': [\"I'm sorry about all of Sheldon's interruptions.\\nHe can be a bit of an eccentric.\",\n","   'If by \"eccentric\" you mean passive-aggressive East Texas blowhole, I agree.'],\n","  'context_speakers': ['LEONARD', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Provided he has already read and is familiar with the reestablishment of the DC multiverse.',\n","  'speaker': 'SHELDON',\n","  'context': [\"Everybody's staring at me.\",\n","   \"Don't worry, they're more scared of you than you are of them.\",\n","   'Unlikely.',\n","   'Here, what about this one for my nephew?',\n","   'A superb choice.',\n","   'Oh, great.'],\n","  'context_speakers': ['PENNY',\n","   'LEONARD',\n","   'PENNY',\n","   'PENNY',\n","   'SHELDON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'No, I mean, we can always go back to being friends.',\n","  'speaker': 'PENNY',\n","  'context': [\"like Mrs. Riley' s chicken.\",\n","   \"What's Mrs. Riley's chicken?\",\n","   'A chicken that was owned by Mrs. Riley.',\n","   'Okay, forget the chicken.',\n","   'Well, I wish I could.',\n","   'No, no. You may be right about me and Leonard.',\n","   \"Of course I'm right What are the odds I'd be wrong twice in one week?\"],\n","  'context_speakers': ['SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'PENNY',\n","   'SHELDON',\n","   'PENNY',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Seems like an awful lot of trouble to go through for intercourse.',\n","  'speaker': 'SHELDON',\n","  'context': [\"Am I correct in assuming that your attempt to be accepted by Penny's peers is based on your desire to ensure your continuing mating privileges with her?\",\n","   \"Well, I wouldn't put it exactly that way.\",\n","   'How would you put it?',\n","   'Yeah, okay, like you said.'],\n","  'context_speakers': ['SHELDON', 'LEONARD', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Not a great movie, but look at that beautiful desert.',\n","  'speaker': 'RAJ',\n","  'context': ['We did it. We have the west coast feed of hbo.',\n","   'Ooh,real sex.',\n","   \"Big deal. Every time I watch that show, it's old women putting condoms on cucumbers.\",\n","   \"All right,let's see what's on the east coast feed.\",\n","   'Dune.'],\n","  'context_speakers': ['RAJ', 'HOWARD', 'RAJ', 'LEONARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Raj, you should've seen Leonard's experiment. The interference pattern was so cool when the electron beam was on.\",\n","  'speaker': 'BERNADETTE',\n","  'context': ['Should I give him another chance?',\n","   \"It's up to you. He didn't call your experiment stupid.\",\n","   'Come here, tushy face.',\n","   '\"Tushy face. \" Raj, you should\\'ve seen Leonard\\'s experiment.'],\n","  'context_speakers': ['BERNADETTE', 'LEONARD', 'BERNADETTE', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I was going to ask Leonard to do it, but he seemed a bit emotionally unstable and you don't want someone like that operating heavy machinery.\",\n","  'speaker': 'PERSON',\n","  'context': ['Thank you for driving me back to my hotel.',\n","   \"Oh,it's not a problem.\"],\n","  'context_speakers': ['PERSON', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Is it possible that she was going for your cheek and you moved and she accidentally got lip? That happens with me and my mom all the time.',\n","  'speaker': 'HOWARD',\n","  'context': ['to bring us all with you to the year 7010 where we are transported to work at the thinkatorium by telepathically controlled flying dolphins?',\n","   'No, Penny kissed me.'],\n","  'context_speakers': ['SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Well, I'm just glad you finally found yourself a keeper.\",\n","  'speaker': 'PERSON',\n","  'context': [\"and pump it into his mom's Camry.\",\n","   \"Yeah, that's great. It's a funny story. Moving on.\",\n","   'And Donnie was a rocket scientist compared to that boy who wanted to get beer pong into the olympics.',\n","   'What was his name, sweetheart?',\n","   \"Curtis, and I'm pretty sure he was joking.\",\n","   \"I don't know. That petition looked real to me.\",\n","   'Yeah, okay, I think this ends the ex-boyfriend portion of our evening.'],\n","  'context_speakers': ['PERSON',\n","   'PENNY',\n","   'PERSON',\n","   'PERSON',\n","   'PENNY',\n","   'PERSON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Obviously, we're no longer a Justice League. We have no choice but to switch to our Muppet Baby costumes.\",\n","  'speaker': 'SHELDON',\n","  'context': ['I may have failed.',\n","   'Okay, I guess we just go without a Wonder Woman.',\n","   \"I don't want to go without Penny.\",\n","   \"Oh, great. No Superman, no Wonder Woman? All we've got is a skinny Flash, an Indian Aquaman, a nearsighted Green Lantern, and a teeny, tiny Dark Knight.\"],\n","  'context_speakers': ['SHELDON', 'HOWARD', 'PERSON', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Certainly. I'd like to raise two points. Number one, I think they <i>are</i> talking about penises. And number two, these mimosas are kicking my little brown ass.\",\n","  'speaker': 'RAJ',\n","  'context': ['Who could get <i>all</i> the giant missiles she wants.',\n","   \"Are we talking about women wanting penises? Because I'd like to weigh in.\",\n","   'Dr. Koothrappali, would you care to join the conversation?'],\n","  'context_speakers': ['BERNADETTE', 'AMY', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I'm not saying that you're not fun. You're the most fun person I know.\",\n","  'speaker': 'AMY',\n","  'context': ['Your flaring nostrils indicate otherwise.',\n","   'Cluck-cluck-cluck.',\n","   \"Sheldon, look at me.\\nI think it's time to face the fact that Leonard is the nucleus of your social group.\\nWhere he goes, the group goes.\",\n","   'Leonard the nucleus? That makes no sense.',\n","   \"I'm the whimsical elf that everyone looks to for a good time.\"],\n","  'context_speakers': ['AMY', 'SHELDON', 'AMY', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'All right, so technically it\\'s not a dinner date. I suppose you could call it a, uh, dinfast date. But if you did, you\\'d open yourself to peer-based mocking, such as, \"Hey, Leonard, how was your dinfast with Priya last night?\"',\n","  'speaker': 'SHELDON',\n","  'context': ['What are you doing?',\n","   \"Oh, uh, Priya's calling in a few minutes on Skype, and we are gonna have a dinner date.\",\n","   \"It's 8:00 in the morning in Mumbai. \\nHow can she have dinner?\",\n","   'Fine, whatever. \\nPriya will be having breakfast.'],\n","  'context_speakers': ['SHELDON', 'LEONARD', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Because you don't have a girlfriend? Well, good Lord, if that becomes a reason not to play Dungeons & Dragons, this game's in serious trouble.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Your movies and your TV shows promised streets paved with beautiful blonde women with big bazongas.',\n","   \"Eat another pie, you'll have your own bazongas.\",\n","   \"That's cruel.\",\n","   'You know it goes straight to my hips.',\n","   'Gentlemen, please focus.',\n","   \"You're facing a fire-breathing dragon.\",\n","   \"I don't know if I want to play anymore.\"],\n","  'context_speakers': ['RAJ',\n","   'HOWARD',\n","   'RAJ',\n","   'RAJ',\n","   'SHELDON',\n","   'SHELDON',\n","   'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'No, but maybe she wants a man with a pocket watch.',\n","  'speaker': 'SHELDON',\n","  'context': [\"I don't think there's anything in this jewelry store that Amy would appreciate more than the humidifier we were just looking at at Sears.\",\n","   'Oh, my God, now I know what I sound like to you when I say stupid stuff.',\n","   'Ooh, a pocket watch.',\n","   \"Okay, I don't think Amy wants a pocket watch.\"],\n","  'context_speakers': ['SHELDON', 'PENNY', 'SHELDON', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'The original title was \"A Rederivation of Maxwell\\'s Equations Regarding Electromagnetism\"\" I dumbed it down because some of the more religious people in town were starting to say I was a witch.',\n","  'speaker': 'SHELDON',\n","  'context': ['This will interest you. And, Penny, feel free to paint your nails.',\n","   'What do you got?',\n","   'I believe Alex may have unearthed the grade school science project that could be my ticket to the Nobel Prize.',\n","   'Behold.',\n","   '\"Magnets: What Do They Stick To?\" If the answer is \"metal,\" it\\'s not exactly groundbreaking.'],\n","  'context_speakers': ['SHELDON', 'LEONARD', 'SHELDON', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Trust me, you're not Jewish. That's how they get you.\",\n","  'speaker': 'HOWARD',\n","  'context': ['Rajesh!\\nYou up?\\nYou ready for breakfast?',\n","   'Oh, boy. Breakfast.',\n","   'Okay, Raj, listen to me.\\nYou need to get out of there.',\n","   'But I have a cream sherry hangover and I smell blintzes.',\n","   \"If you don't leave now, she'll use food and guilt to keep you there for the rest of your life.\",\n","   'Oh, Howard, stop.'],\n","  'context_speakers': ['PERSON', 'RAJ', 'HOWARD', 'RAJ', 'HOWARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Not as relieved as I'm about to be. It's a brave new world, little lady.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Really?',\n","   'Yes.',\n","   'Huh. Yeah, I always took it literally.',\n","   \"That's why I have never once moved my bowels in this or any restaurant. \\nHmm.\",\n","   \"I'm relieved that you don't have a problem with us working together.\"],\n","  'context_speakers': ['SHELDON', 'AMY', 'SHELDON', 'SHELDON', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'That woman looks exactly like the pictures of Princess Panchali in the book. How often does one see a beloved fictional character come to life?',\n","  'speaker': 'SHELDON',\n","  'context': ['When I was a little boy and got sick, which was most of the time, my mother would read it to me.',\n","   \"It's about an Indian princess who befriends a monkey who was mocked by all the other monkeys because he was different.\",\n","   'For some reason, I related to it quite strongly.',\n","   'I know the reason.',\n","   'We all know the reason.',\n","   'Sheldon, what are you getting at?'],\n","  'context_speakers': ['SHELDON',\n","   'SHELDON',\n","   'SHELDON',\n","   'PENNY',\n","   'LEONARD',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Technically, yes. But, if you'll notice... It's reversible!\",\n","  'speaker': 'LEONARD',\n","  'context': [\"Because it's not a costume, it's a flight suit.\",\n","   'Ok, alright, moving on.',\n","   'Oh, wow. A paisley shirt.',\n","   'It goes with my corduroy suit.',\n","   'If you mean it should end up in the same place, then I agree.',\n","   'Is this your only tie?'],\n","  'context_speakers': ['LEONARD',\n","   'PENNY',\n","   'PENNY',\n","   'LEONARD',\n","   'PENNY',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'What else? Sell it on eBay as \"slightly used.\"',\n","  'speaker': 'RAJ',\n","  'context': [\"Or maybe she just doesn't want to talk.\",\n","   'Look, I found an iPod!',\n","   \"It's smashed beyond repair. What are you gonna do with it?\"],\n","  'context_speakers': ['SHELDON', 'RAJ', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I don't know. It just died. I'm getting a bunch of static.\",\n","  'speaker': 'PENNY',\n","  'context': ['Hey, guys, I need to use your TV.',\n","   \"What's wrong with your TV?\"],\n","  'context_speakers': ['PENNY', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'The big deal is that nobody touches food on my plate.',\n","  'speaker': 'SHELDON',\n","  'context': ['Who touched my…',\n","   'Penny! Penny did it.',\n","   'Why would you do that?',\n","   \"I don't know. I was hungry?\\nWhat's the big deal?\"],\n","  'context_speakers': ['SHELDON', 'HOWARD', 'SHELDON', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'What do you think I did? I discreetly slipped off the other earring, put it in my pocket and then got the hell out of there!',\n","  'speaker': 'PERSON',\n","  'context': [\"So, we're all standing around looking at the post-op X ray and there it is, clear as day, right in the guy's chest cavity...one of my earrings.\",\n","   'Oh, my God, what did you do?'],\n","  'context_speakers': ['PERSON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'My aunt Marion gave them to me for my 12th birthday.',\n","  'speaker': 'SHELDON',\n","  'context': ['What are you doing here?',\n","   \"Hang on. 130 over 80. It's a little high.\",\n","   'We can attribute that to the stress of sneaking past the security desk.',\n","   'Where did you get the stethoscope and the blood pressure cuff?'],\n","  'context_speakers': ['PERSON', 'SHELDON', 'SHELDON', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"No, I'm behind on my Wiki-reading. I'm kind of on a John Grisham kick right now.\",\n","  'speaker': 'RAJ',\n","  'context': [\"I think we're fitting in quite nicely.\",\n","   \"It'd help if you weren't drinking light beer.\",\n","   \"What's so gothic about vodka and cranberry juice?\",\n","   'Hello? It looks like blood.',\n","   'Did you even read the \"Wiki how\" link I sent you on being Goth?'],\n","  'context_speakers': ['RAJ', 'HOWARD', 'RAJ', 'HOWARD', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Oh, you know, Goth stuff. Goth magazines, Goth music.',\n","  'speaker': 'HOWARD',\n","  'context': ['So, what do you guys do?'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Oh, no, no, I said I'm inviting you to come with me to Stan Lee's house.\",\n","  'speaker': 'PENNY',\n","  'context': [\"Hi. I'm Penny. This is my friend, Sheldon.\",\n","   \"We're not friends at the moment. Depends on how this goes.\",\n","   'Right, right. Anyway, Sheldon here is a huge fan of yours, and he was supposed to meet you the other day at the comic book store, but he kind of ended up in jail.',\n","   \"I see. And you thought you'd just come over to my house uninvited?\",\n","   'You said we were invited.'],\n","  'context_speakers': ['PENNY', 'SHELDON', 'PENNY', 'PERSON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I read a study online that walking after a meal not only aids in digestion, but increases serotonin, and you know me, if there's one thing I like more than a refreshing beverage, it's serotonin. Bye-bye.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Ah…I do so love beverages.',\n","   \"Now I think I'll take my after-dinner walk.\",\n","   'Since when do you take after-dinner walks?',\n","   'Yeah, since when do you take walks?'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'LEONARD', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'It was the kind of thing that makes it kind of hard to face her now.',\n","  'speaker': 'HOWARD',\n","  'context': ['You never told me what happened between you and Bernadette.',\n","   'I did a stupid thing.',\n","   'Yeah, I guessed that.'],\n","  'context_speakers': ['LEONARD', 'HOWARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Bernadette dared me to tell a dirty story. \"The Miller\\'s Tale\" by Chaucer is the dirtiest story I know. It would have been hidden in sock drawers if people in the 14th century had worn socks.',\n","  'speaker': 'AMY',\n","  'context': ['\"And Absolon hath kist hir nether ye, \"and Nicholas is scalded in the towte.\\nThis tale is doon, and God save al the rowte.\"',\n","   'What the hell was that?'],\n","  'context_speakers': ['AMY', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Well, yeah, sex with you is pretty great. Have you ever tried it?',\n","  'speaker': 'LEONARD',\n","  'context': ['That was a really nice dinner.',\n","   \"I'm glad you asked me out again.\",\n","   'Me, too; I missed you.',\n","   'You see me all the time.',\n","   \"You sure you just don't miss the sex?\"],\n","  'context_speakers': ['PENNY', 'PENNY', 'LEONARD', 'PENNY', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Oh, screw her. She's just lucky we let her hang out with us.\",\n","  'speaker': 'AMY',\n","  'context': ['The painting is a constant reminder that of the three of us, she is the least cool.',\n","   \"Yeah, that's... that's what it is, so...\",\n","   'You have such a good heart.',\n","   'Yeah... I try.',\n","   'Come on.',\n","   'Whoa, whoa, where are we going?',\n","   \"We're going to go put this painting back up in your apartment.\",\n","   'Well, what about Bernadette?'],\n","  'context_speakers': ['AMY',\n","   'PENNY',\n","   'AMY',\n","   'PENNY',\n","   'AMY',\n","   'PENNY',\n","   'AMY',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I am talking to a man in space. If you don't have the (walkie-talkie static) then he might as well be at the Coffee Bean over on Lake Street.\",\n","  'speaker': 'SHELDON',\n","  'context': ['You know, joking, kidding around, hurting feelings.',\n","   'Okay, my turn. \\nLet me talk to him.',\n","   'Go ahead.',\n","   '2311 North Los Robles Avenue, Pasadena, California to International Space Station.\\nCan you read me? Over.',\n","   '(imitates walkie-talkie static) Yes, I read you, Sheldon.',\n","   'Copy that. Over. (imitates walkie-talkie static)',\n","   'What are you doing?'],\n","  'context_speakers': ['HOWARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'HOWARD',\n","   'SHELDON',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Of course, it is... it's been in the news. And it's a very famous boson.\",\n","  'speaker': 'PENNY',\n","  'context': ['There are some charts in the back where I kept track of shape, color and consistency.',\n","   'Disgusting.',\n","   \"No, what's disgusting is he's still keeping track.\",\n","   'Why do you need all this stuff...?',\n","   'No, no, no, no... Sorry!',\n","   'I am glad you asked.',\n","   'Yeah, are you familiar with the Higgs boson?'],\n","  'context_speakers': ['SHELDON',\n","   'PENNY',\n","   'LEONARD',\n","   'PENNY',\n","   'LEONARD',\n","   'SHELDON',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"We've kind of been involved in a five-year experiment.\",\n","  'speaker': 'PENNY',\n","  'context': ['in town were starting to say I was a witch.',\n","   'Sheldon thinks the approach in this paper might change the way we calculate ferromagnetic hysteresis.',\n","   \"Oh, it's about time. I hated the old way.\",\n","   \"Hi. I'm Penny.\",\n","   'Alex.\\nUh, do you work with Dr. Hofstadter?',\n","   'In a way.'],\n","  'context_speakers': ['SHELDON',\n","   'PERSON',\n","   'PENNY',\n","   'PENNY',\n","   'PERSON',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'My limbic system wants to take your pants off.',\n","  'speaker': 'PENNY',\n","  'context': ['Uh, do you work with Dr. Hofstadter?',\n","   'In a way.',\n","   \"We've kind of been involved in a five-year experiment.\",\n","   \"Oh. Well, you're lucky. He seems very talented.\\nAnd I'm sure a lot of people want to work with him.\",\n","   \"Well, a lot of people can't.\",\n","   'Uh, hey.',\n","   'Where are we going?'],\n","  'context_speakers': ['PERSON',\n","   'PENNY',\n","   'PENNY',\n","   'PERSON',\n","   'PENNY',\n","   'PENNY',\n","   'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'I did. The guy at the liquor store said this stuff tastes great in coffee.',\n","  'speaker': 'RAJ',\n","  'context': ['Do you really think you should be drinking right now?',\n","   'How else am I supposed to talk to the Human Resources lady?',\n","   \"I don't know. Seek professional help?\"],\n","  'context_speakers': ['HOWARD', 'RAJ', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"No, just me, her and a couple of the big wigs. It's part of a bonus she got.\",\n","  'speaker': 'HOWARD',\n","  'context': ['Here you go.',\n","   'Great. Come on in.',\n","   \"What, you don't say thank you?\",\n","   \"It's my suitcase.\\nI lent it to you two years ago.\",\n","   'Well, then, I should tell you I broke the wheel and the handle.',\n","   \"So, is everyone from Bernadette's company going to Vegas?\"],\n","  'context_speakers': ['RAJ', 'HOWARD', 'RAJ', 'HOWARD', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Go out with me on one date, and I promise you, you'll see.\",\n","  'speaker': 'RAJ',\n","  'context': ['There are many things seriously wrong with me.',\n","   'And not quirks, either.\\nLike... diagnosable psychological problems.',\n","   'Maybe brain damage.',\n","   \"Well, how do I know you're not just saying that?\"],\n","  'context_speakers': ['RAJ', 'RAJ', 'RAJ', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Okay, sit down. Listen, Emily, what happened between me and Raj was a long time ago. It was,',\n","  'speaker': 'PENNY',\n","  'context': ['Uh, listen, Penny, before you start, I just want to apologize for being so rude the other day in my office.',\n","   'No, no. No need to apologize.',\n","   'Excuse me, can I say something?',\n","   'I just love both you guys, and I want you to get along.',\n","   'And when I say love\" I meant as a friend.\\nAnd, uh, way too soon, right?'],\n","  'context_speakers': ['PERSON', 'PENNY', 'RAJ', 'RAJ', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Well, he said it was a tribble. It could be a toupee, but either way, it's pretty cool.\",\n","  'speaker': 'PERSON',\n","  'context': ['Ooh.\\nHe was so happy after he passed it, he gave me a Terminator.',\n","   'That is amazing.\\nHow long have you been collecting?',\n","   \"Ever since I was a kid, but, uh, I didn't really get serious until William Shatner's bladder infection.\",\n","   \"What'd you get? What'd you get?\"],\n","  'context_speakers': ['PERSON', 'LEONARD', 'PERSON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Do you think Penny will come here and take care of us?',\n","  'speaker': 'SHELDON',\n","  'context': ['I want grilled cheese.'],\n","  'context_speakers': ['SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Would it work if I just punched you in the face?',\n","  'speaker': 'LEONARD',\n","  'context': ['it will establish you as the \"alpha\" male.',\n","   'You see... when a female witnesses an exhibition of physical domination, she produces the hormone oxytocin.',\n","   'If the two of you then engage in intercourse, this will create the biochemical reaction in the brain which lay people naively interpret as \"falling in love.\"',\n","   'Huh'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Excuse me, but what about me? Why don't I get a shot?\",\n","  'speaker': 'LEONARD',\n","  'context': [\"Unlike you, I can actually talk to women when I'm sober.\",\n","   'You fail to take into account that even mute, I am foreign and exotic, while you, on the other hand, are frail and pasty.',\n","   'You know the old saying, \"Pasty and frail never fail.\"'],\n","  'context_speakers': ['HOWARD', 'RAJ', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"You know this can go on all night. Why don't you just come with us?\",\n","  'speaker': 'LEONARD',\n","  'context': ['Or pick out his career for him without knowing his aptitude. Or pick out a new breakfast cereal without knowing his fiber requirements.',\n","   'Or his feelings about little marshmallows.',\n","   'Spider-Man. Get him Spider-Man.',\n","   'Amazing Spider-Man, Ultimate Spider-Man, Spectacular Spider-Man, The Marvelous Adventures of Spider-Man, Spider-Man 2099?',\n","   'You know this can go on all night.'],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'PENNY', 'SHELDON', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"You don't really believe in that superstition, do you?\",\n","  'speaker': 'SHELDON',\n","  'context': ['This thing has to hold up against a hearty potato-based diet.',\n","   'I feel terrible',\n","   \"Maybe if you were helping, you'd feel better about yourself.\",\n","   \"I deliberately tried to sabotage Stuart's date with Penny.\",\n","   'Of course you feel terrible. You completely screwed up your karma.'],\n","  'context_speakers': ['HOWARD', 'LEONARD', 'HOWARD', 'LEONARD', 'RAJ'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'What did you mean when you said you were going to miss me?',\n","  'speaker': 'LEONARD',\n","  'context': ['What time is it?',\n","   \"It's 7:00 a.m. I'm sorry it's early, but we're leaving soon, and I needed to talk to you.\"],\n","  'context_speakers': ['PENNY', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Don't you think looking for a new city to live in is a bit of an overreaction?\",\n","  'speaker': 'LEONARD',\n","  'context': ['My new computer came with Windows 7.',\n","   'Windows 7 is much more use r- friendly than Windows Vista.',\n","   \"I don't like that.\"],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Why don't you keep that stuff in the bathroom?\",\n","  'speaker': 'HOWARD',\n","  'context': [\"That's a great idea. I'm going to get him to sign a Batman as well.\",\n","   'What is it about the word \"unique\" you don\\'t understand?',\n","   'Damn, paper cut. Nothing worse than a paper cut.',\n","   \"Well, obviously you don't remember your circumcision.\",\n","   \"There's Neosporin and Band-Aids in my top desk drawer.\"],\n","  'context_speakers': ['RAJ', 'SHELDON', 'HOWARD', 'RAJ', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"So what, you're just gonna walk up and down the stairs?\",\n","  'speaker': 'HOWARD',\n","  'context': [\"I've missed you.\",\n","   'All right, say hello to your mother for me.',\n","   'Okay.',\n","   'What?',\n","   'You said you were going for a walk.',\n","   \"I didn't say outside.\"],\n","  'context_speakers': ['RAJ',\n","   'SHELDON',\n","   'HOWARD',\n","   'SHELDON',\n","   'HOWARD',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'I just have a question. Does Bernadette ever talk about me?',\n","  'speaker': 'HOWARD',\n","  'context': ['Would you have opened the door if you knew it was me?',\n","   'Not since I found out the teddy bear you gave me had a webcam in it.'],\n","  'context_speakers': ['HOWARD', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'I understand. Your loyalty is admirable. But what if I were to up the ante?',\n","  'speaker': 'SHELDON',\n","  'context': ['Why not \"World\\'s Greatest Engineer\"?',\n","   \"I'm sorry.\\nI typed that into the label maker, but I just couldn't press enter. Now down to business.\",\n","   'What would it take for you to abandon Leonard and join me in a rival company',\n","   'Unless, of course, the mugs are sufficient, in which case, welcome aboard.',\n","   \"We're not quitting on Leonard.\"],\n","  'context_speakers': ['HOWARD', 'SHELDON', 'SHELDON', 'SHELDON', 'HOWARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I'm sorry, Sheldon bought a loom and learned how to weave?\",\n","  'speaker': 'PERSON3',\n","  'context': [\"Baby, you're not that kind\",\n","   \"Don't go breaking my heart\",\n","   'Anybody do \"Walking on Sunshine\" yet?',\n","   'No',\n","   'Dibs.'],\n","  'context_speakers': ['PERSON1', 'PERSON1', 'PERSON2', 'SHELDON', 'PERSON2'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'No. You mean this was all a ruse? Oh, how could I be so stu...',\n","  'speaker': 'SHELDON',\n","  'context': [\"Oh, no! I can't!\",\n","   \"Hurry! \\nWe're running out of time!\",\n","   '- Okay.',\n","   'Just do it!',\n","   'Oh, God!\\nOne... two...\\nthree!',\n","   'Trick or treat, bubbeleh.',\n","   'What?\\n(laughing)'],\n","  'context_speakers': ['SHELDON',\n","   'BERNADETTE',\n","   'SHELDON',\n","   'BERNADETTE',\n","   'SHELDON',\n","   'HOWARD',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Yeah, that was five weeks ago. How much longer is he gonna milk that cow?',\n","  'speaker': 'SHELDON',\n","  'context': ['There it is again.\\nDo you kiss your mother with that mouth?',\n","   \"Oh, well, I'm sorry for your loss.\\n- Oh! - Good night, sir.\",\n","   \"Unbelievable.\\nHe says Wolowitz deserves my spot because he's a high-profile asset to the university.\",\n","   \"Well, he's not wrong.\\nHoward did go to the International Space Station.\"],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"- No, no, no. I'm really good at this. All right, give me a hint. Does it involve difficulty initiating a stream of anything?\",\n","  'speaker': 'SHELDON',\n","  'context': ['Dr. Jeffries, hello again. Leonard Hofstadter.',\n","   'Oh, oh, right, I remember your-your girlfriend.',\n","   'Is-Is she... is she here?',\n","   \"No, she's not.\",\n","   'So, what prescription are you getting filled?',\n","   '- Sheldon.',\n","   \"No. Wait, I want to guess. Don't tell me.\",\n","   \"I wasn't going to tell you.\"],\n","  'context_speakers': ['LEONARD',\n","   'PERSON',\n","   'PERSON',\n","   'LEONARD',\n","   'SHELDON',\n","   'LEONARD',\n","   'SHELDON',\n","   'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'What do you think I was gonna do at home?',\n","  'speaker': 'PERSON',\n","  'context': ['Koothrappali murder mystery dinner.',\n","   \"I'm leaving.\",\n","   \"You can't leave. \\nYou're a suspect in the mysterious murder of Stuart Bloom.\",\n","   \"(dramatic music plays) I didn't know his last name was Bloom.\",\n","   \"Yeah, it's Bloom.\",\n","   'Are you really gonna lie on the floor and pretend to be dead all night?'],\n","  'context_speakers': ['RAJ', 'AMY', 'RAJ', 'BERNADETTE', 'AMY', 'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'And of those few months, how long have you been a demented sex pervert?',\n","  'speaker': 'SHELDON',\n","  'context': [\"Well, I'm sure that, uh, you...\\nhave a lot of questions.\",\n","   'Who was he?',\n","   'His name is Ron.\\nI met him at my prayer group.',\n","   'How long have you been involved with him?',\n","   'A few months.'],\n","  'context_speakers': ['PERSON', 'SHELDON', 'PERSON', 'SHELDON', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I'm sorry, I am not going back to the Renaissance fair.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"That's enough.\",\n","   'Call me.',\n","   'Right.',\n","   'Okay, good night.',\n","   'What?',\n","   'Had a great time. Ciao.'],\n","  'context_speakers': ['PERSON1',\n","   'PERSON1',\n","   'LEONARD',\n","   'PENNY',\n","   'PERSON2',\n","   'PENNY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'You were very clear. As was everyone else at the table.',\n","  'speaker': 'SHELDON',\n","  'context': ['Leonard, this is Tom.',\n","   'Hi, Tom. Sheldon?',\n","   \"Didn't I explain to you about your little mistake in the cafeteria?\"],\n","  'context_speakers': ['SHELDON', 'LEONARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Thank you so much for giving me the opportunity to help.',\n","  'speaker': 'LEONARD',\n","  'context': ['Thanks so much for helping me.'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"There's no hurry. Tell them more about their secret love for each other.\",\n","  'speaker': 'LEONARD',\n","  'context': ['Say what?',\n","   \"That's basically what I just said.\",\n","   'You brought your husband to work. You know the rules.',\n","   'Mesa thinking yousa looking pretty sad now too, betcha, betcha.',\n","   \"Leonard, it's 1:00. Weren't you going to show me your laboratory at 1:00?\"],\n","  'context_speakers': ['HOWARD', 'HOWARD', 'LEONARD', 'LEONARD', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Right. That was so good! Sweety and meaty at the same time.',\n","  'speaker': 'RAJ',\n","  'context': ['Anything?',\n","   'No. that was the last pudding cup.',\n","   'No! What about slim jims',\n","   'Thats what he used to eat his pudding, remember?'],\n","  'context_speakers': ['HOWARD', 'LEONARD', 'HOWARD', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'All right, all right. He can buy his own desk.',\n","  'speaker': 'SHELDON',\n","  'context': ['and a super executive ant farm with glow-in-the-dark sand?',\n","   '- What if he buys his own desk?',\n","   'Yeah, what if I buy my own desk?',\n","   \"- That's ridiculous.\",\n","   '- Why?',\n","   'Because…',\n","   'Yes?',\n","   \"It's my office.\"],\n","  'context_speakers': ['SHELDON',\n","   'HOWARD',\n","   'RAJ',\n","   'SHELDON',\n","   'RAJ',\n","   'SHELDON',\n","   'RAJ',\n","   'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"He does it all the time. He's a cornucopia of social awkwardness.\",\n","  'speaker': 'SHELDON',\n","  'context': ['Right.',\n","   \"That's not true of all bars. Juice bars, for example.\",\n","   'Oh! Good point, Amy.\\nYeah, building on your premise, Leonard could frequent sushi bars, salad bars, oyster bars, the Apple Genius Bar... What are you doing?',\n","   \"Keep going. I'm listening.\",\n","   'That was rude.'],\n","  'context_speakers': ['LEONARD', 'AMY', 'SHELDON', 'LEONARD', 'AMY'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'So, this spring, I get to go to the International Space Station. Oh, my word, a trip to the heavens.',\n","  'speaker': 'HOWARD',\n","  'context': [\"Now I'm going out tonight. Would it be crazy to ask you to look at the outfit I'm going to wear?\",\n","   \"Oh, not crazy at all.\\nAnd don't beat yourself up.\",\n","   'When I was your age, you could have me for a car ride and a bottle of strawberry wine.',\n","   \"That will not be in this week's e-mail blast.\"],\n","  'context_speakers': ['PENNY', 'PERSON', 'PERSON', 'SHELDON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I'm sorry, it's the bad boy attitude that comes with this hair.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"You know, Sheldon, I used to cut my brother's hair. \\nI could do it for you.\",\n","   'Penny, I know you mean well, offering the skills of the hill folk, but, um,',\n","   \"here in town, we don't churn our own butter, we don't make dresses out of gunny sacks, and, uh, we sure as shootin' don't get our hair cut by bottle-blonde...\",\n","   'Sheldon, be nice.'],\n","  'context_speakers': ['PENNY', 'SHELDON', 'SHELDON', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"I don't need a reason-- it's my house.\",\n","  'speaker': 'HOWARD',\n","  'context': [\"It's true-- you're both gorgeous.\",\n","   'Kind of says something about the man who could bed you both.',\n","   \"You get why I've been alone most of my adult life.\",\n","   'So, what brings you by? Well,'],\n","  'context_speakers': ['RAJ', 'RAJ', 'RAJ', 'PERSON'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': \"Oh, you're hardly out of the woods, no. You still run the risk of infection, a blood clot, the possibility that an inattentive surgeon let a barn spider lay eggs in your nose.\",\n","  'speaker': 'SHELDON',\n","  'context': [\"Wow. I don't know which hurts worse-- my nose or my heart.\",\n","   \"Well, I'm done speaking to you.\",\n","   \"Don't be like that. You two need to talk this out.\",\n","   \"Yeah, 'cause you sound really funny.\",\n","   \"Sheldon, I'm sorry I didn't tell you about the surgery, but you were worried about nothing.\"],\n","  'context_speakers': ['SHELDON', 'SHELDON', 'AMY', 'PENNY', 'LEONARD'],\n","  'show': 'BBT',\n","  'sarcasm': False},\n"," {'utterance': 'Why? You never look down in the shower?  Oh please I am allowed one joke in the monkey is penis genre.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"This is so exciting! I haven't seen my monkey in almost a year\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'No, no the actual cartoon character .. Of course the baloon.',\n","  'speaker': 'CHANDLER',\n","  'context': ['The most unbelievable thing has happened. Underdog has gotten away.'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Spanish midgets wrestling Julie! Okay yes I see how they got there.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Man, I sure miss Julie'],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I am so confused as to what we have been doing so far.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Thirty seconds all the questions we can answer',\n","   'You guys are dead, I am so good in lightning rounds.',\n","   'I majored in lightning rounds, alright! We are gonna destroy you!',\n","   'Wanna bet?'],\n","  'context_speakers': ['ROSS', 'MONICA', 'CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'We think it was trying to spell out monkey.',\n","  'speaker': 'CHANDLER',\n","  'context': ['He also found an M and an O.'],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Hey, that's Joincidence with a C\",\n","  'speaker': 'CHANDLER',\n","  'context': ['You know what I just realized? Joker is poker with a J. Coincidence?'],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Your makeup?',\n","  'speaker': 'ROSS',\n","  'context': [\"Are you referring to my man's bag? At first it I thought it just look good. But its practical too, check it out! Its got compartments for all your stuff. Your wallet, your key, your addressbook.\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Great! Now we can do laundry all night. All night laundry!',\n","  'speaker': 'CHANDLER',\n","  'context': ['Actually, I was gonna do some laundry. Chandler, do you wanna do it with me?',\n","   'Sure I will do it with you',\n","   \"You don't mind do you, that would really help me a lot Thanks.\",\n","   \"I don't think I have enough quarters\",\n","   'I have quarters!'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'RACHEL', 'MONICA', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"I can see it from right here. It'll cost you one husband.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Can you give me a closer look so that I can see the price?'],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I am not freaking out, why would I be freaking out? A woman named Hildy called and said we will get married, but that happens everyday',\n","  'speaker': 'CHANDLER',\n","  'context': ['I am so sorry, please stop freaking out.'],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Oh! Satan's minions at work again?\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"And there's five hundred extra dollars in my account.\"],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'This is so good, that I am gonna enjoy it in the balcony. I can enjoy my view, whilst I enjoy my desert.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Mmm', 'Mmm'],\n","  'context_speakers': ['CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'No, rabies.',\n","  'speaker': 'CHANDLER',\n","  'context': ['What what?', 'Hey, shaving?'],\n","  'context_speakers': ['CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Excellent hole, Joe.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Enhh',\n","   \"I've got canned goods.\",\n","   'Hey, you guys! Take a look at this!  Check this baby out, dug me a hole!'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'ThatÕs a good one too Pheebs, now all that I have to do is find a plane load of people whose resolution is to plummet to their deaths',\n","  'speaker': 'CHANDLER',\n","  'context': [\"That's a good one. Mine is to pilot a commercial jet.\"],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'These people find out that a spelling game is going over there, there will be nobody left at our party.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Could guys please try to keep It down? We are trying to start a tournament.'],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"I mean, you and I we're just, y'know, we're nothing, we're goofin' around.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Umm, listen there's something I think you should know.\",\n","   'Oh, is this about you-you dating the nurse? Yeah, Joey already told me, and I am so-so fine.'],\n","  'context_speakers': ['MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'And I think my testicles may be in here too',\n","  'speaker': 'CHANDLER',\n","  'context': ['Do you guys have any exta ribbon?',\n","   'Oh yeah sure, what do you need? We got lace, satin, sateen, rafia, gingam , felt.'],\n","  'context_speakers': ['RACHEL', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'No thanks Josephine!',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey Chandler!, come on in we are knitting pot holders.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Even if people are on it!\"',\n","  'speaker': 'RACHEL',\n","  'context': ['You know what else I\\'m not gonna miss? \"I\\'m Monica. I wash the toilet 17 times a day.'],\n","  'context_speakers': ['RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"I guess you'd know that, being one of the two of us, though, right?\",\n","  'speaker': 'RACHEL',\n","  'context': [\"So. I just thought the two of us should hang out for a bit. I mean, you know, we've never really talked.\"],\n","  'context_speakers': ['RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'That monkey has got a Ross on its ass!',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey!'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Well, I'd gotten Judy pregnant. I still don't know how that happened.\",\n","  'speaker': 'PERSON',\n","  'context': ['I remember when we first got engaged.',\n","   \"Oh, I don't think I ever heard that story.\"],\n","  'context_speakers': ['PERSON', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Oh I left them in my bulldozer .. I don't have tools!\",\n","  'speaker': 'CHANDLER',\n","  'context': ['It isnt working',\n","   'We have to unscrew the chain',\n","   \"Hurry, I can't feel my ears\",\n","   'Can you ever feel your ears?',\n","   'Interesting'],\n","  'context_speakers': ['JOEY', 'MONICA', 'JOEY', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"You know once you are inside, you don't have to knock anymore.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey', 'Hey'],\n","  'context_speakers': ['JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Janice? Janice?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Will you get me a white zinfandel and glass of red for Janice.'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh hurry up, I want to sign that.',\n","  'speaker': 'CHANDLER',\n","  'context': ['But that reminds me, sign this.',\n","   'What is it?',\n","   \"Its nothing, just that you can't sue the city incase you scrap your knee or your know get your head blown off.\"],\n","  'context_speakers': ['PERSON', 'ROSS', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Okay, but if you are not coming back soon, there is pretty much nothing I can do about it',\n","  'speaker': 'CHANDLER',\n","  'context': ['Saucy.',\n","   \"I'll be back in two minutes.\",\n","   'You are going to leave me like this',\n","   \"Knowing you are here, waiting for me, it is kind of exciting isn't it?\"],\n","  'context_speakers': ['CHANDLER', 'PERSON', 'CHANDLER', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'No you are not Chandler, we still love you Chandler',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Okay you are right, I am in love with Joey's girlfriend.\",\n","   'What? How is that possible? You barely know her.',\n","   \"I don't know. I can't . I just, I can't get her out of my head you know. I am a very bad person. I am a horrible person.\"],\n","  'context_speakers': ['CHANDLER', 'PHOEBE', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I guess you would know that or else we would be in a predicament room',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Listen it's kind of an emergency\"],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Really? Struck down in the prime of life.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Chandler, I think I am sick'],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Dear god, this parachute is a knapsack',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Why wouldn't I be married when I am forty?\",\n","   'Oh no no no, I just meant hypothetically',\n","   \"Ok , hypothetically, why won't I be married when I am forty?\",\n","   'No',\n","   'No, what is it? Is there something fundamentally unmarriable about me?'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'MONICA', 'CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh, uh, no thanks. I just had an M&M.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey, you guys the water is great. You should really go in!'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh yeah ok, including the waffles last week, you now owe me, seventeen zillion dollars.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Hey, Chan. Can you help me out here, I promise I'll pay you back.\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'And Hornswoggle? What? Are you dating a character from \"Fraggle Rock\"? Ahh.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I can't take it any more! So you win, okay? Pheebs? Flying a jet? Better make it a spaceship so that you can get back to your home planet! And Ross, phone call for you today. Tom Jones, he wants his pants back!\"],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Let me think! Let me think! Oh, I don't care.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['The wheel or pick a Google card.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Crazy bitch!',\n","  'speaker': 'CHANDLER',\n","  'context': ['Well the tough thing is she really wants to have sex with me.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh tons. I am quite the woodsman.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Okay, lets find these birds.',\n","   'Wait wait, you hear that? There are in the table?',\n","   \"That can't be good\",\n","   'We have to get them out of there',\n","   'how?',\n","   'Maybe we can lure them out, you know of any bird calls?'],\n","  'context_speakers': ['CHANDLER',\n","   'JOEY',\n","   'CHANDLER',\n","   'JOEY',\n","   'CHANDLER',\n","   'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"That's easy for you to say, you weren't almost just killed.\",\n","  'speaker': 'RACHEL',\n","  'context': ['Yes! And he said really mean things that were only partly true.',\n","   \"I'm gonna go down there and teach that guy a lesson.\",\n","   \"Joey, please don't do that. I think it's best that we just forget about it.\"],\n","  'context_speakers': ['RACHEL', 'JOEY', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Uh, Rachel's here, so good luck man, let me know how it works out.\",\n","  'speaker': 'JOEY',\n","  'context': [\"Well, I'm just,if the paste matches the pants, you can make yourself a pair of paste pants and she won't know the difference!\",\n","   'Joey, do you have a minute?',\n","   'Dude, what am I gonna'],\n","  'context_speakers': ['JOEY', 'RACHEL', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Almost! But first, we gotta start.',\n","  'speaker': 'PERSON',\n","  'context': [\"All right, I'm outta here!\",\n","   \"I'm kidding! I'm kidding!\",\n","   'So were done then!'],\n","  'context_speakers': ['RACHEL', 'MONICA', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"All right I'll have one.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"I will have one. Okay, I'm not gonna have one.\",\n","   \"No, no, it's just my tooth.\"],\n","  'context_speakers': ['CHANDLER', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Joey if I go first, I want to be looking for my keys',\n","  'speaker': 'CHANDLER',\n","  'context': ['When their pets pass on, they want them sort of laid out like they are sleeping. But, occassionally theres a person who wants them in a pose like chasing their tail or uh, jumping to catch a frizbee.'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"I can see it from right here. It'll cost you one husband.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Oh wait! Stop! Stop! Stop!',\n","   \"Oh I'm sorry! Do you need a break?\",\n","   'My boots in tan! Hey! Can you get a little closer so I can see the price?'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"So now I have to go so he'll think that I'm totally ok with seeing him!\",\n","  'speaker': 'PHOEBE',\n","  'context': ['And she was like \"Oh, don\\'t worry! I asked him. He\\'s totally ok with seeing you!\".'],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh I see, I thought you just really really liked your new couch.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Look I drew a sketch of how we are gonna do it ok? Rach, that's you, that's the couch\",\n","   \"What's that?\",\n","   \"Oh, that's me.\",\n","   'Wow! Certainly think a lot of yourself.',\n","   \"No, that's my arm.\"],\n","  'context_speakers': ['ROSS', 'RACHEL', 'ROSS', 'RACHEL', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Yeah, I mean, it would be rude to them for us to leave now.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Yeah, I know, it's my birthday. We all should be here.\",\n","   \"So, let's go.\",\n","   'Well maybe, you know, maybe we should stay for one song.'],\n","  'context_speakers': ['ROSS', 'CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh Oh, I am convinced!',\n","  'speaker': 'CHANDLER',\n","  'context': ['What if you could keep the apartment and get the tickets.',\n","   'Done',\n","   'Let me finish, talking about a bet, winner takes all',\n","   'We could end up with nothing.',\n","   'Or you could end up with everything',\n","   'I like that!',\n","   'So what do you say?',\n","   'No',\n","   'Just do it!'],\n","  'context_speakers': ['RACHEL',\n","   'JOEY',\n","   'RACHEL',\n","   'JOEY',\n","   'PHOEBE',\n","   'JOEY',\n","   'MONICA',\n","   'CHANDLER',\n","   'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'If you wanna put a label on it.',\n","  'speaker': 'ROSS',\n","  'context': [\"That's funny. Who are they?\",\n","   'The blond woman is my ex-wife, and the woman touching her is her... close, personal friend.',\n","   \"You mean they're lovers.\"],\n","  'context_speakers': ['PERSON', 'ROSS', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'No, this is just part of a daredevil game that I play called wait until last moment before I burst and die',\n","  'speaker': 'CHANDLER',\n","  'context': ['Someone in there?'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I thought if I littered, that crying Indian might come by and save us.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Chandler, what are you doing? There is a trash can right there!'],\n","  'context_speakers': ['RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Oh boy, I just can't watch. It's too scary!\",\n","  'speaker': 'RACHEL',\n","  'context': ['Oh umm, my big scene is coming up. Big scene coming up.',\n","   'If you said, \"Big lima bean, bubbling up.\" Would she understand the difference?',\n","   'Rach? What are you doing?'],\n","  'context_speakers': ['JOEY', 'CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'The whole thing! Can we go?',\n","  'speaker': 'CHANDLER',\n","  'context': ['You liked it? You really liked it?',\n","   'Oh-ho-ho, yeah!',\n","   'Which part exactly?'],\n","  'context_speakers': ['JOEY', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I can show you an ID if you want?',\n","  'speaker': 'PERSON',\n","  'context': ['Um-hmm. Oh wait! This is Bonnie.',\n","   'Hi!',\n","   \"This is Bonnie?  This is Bonnie?  You're Bonnie?\"],\n","  'context_speakers': ['PHOEBE', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I guess he wants to do a little dance, you know, make a little love. Well pretty much get down to that.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey who is this Casey?',\n","   'Oh some guy she met in the movies',\n","   'Oh really? What does he want with her?'],\n","  'context_speakers': ['ROSS', 'PHOEBE', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh I do. Oh no, wait a minute, I took it out of my shirt when I put it on this morning.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Damn, tail pipe's not hot enough to light this.\",\n","   'Relax ok? I can get this open, anybody got a coat hangar?'],\n","  'context_speakers': ['CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Well, it will probably slow me down at first. But once I am used to the extra weight, I ll get back on track.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Oh no no, that's the beauty of it, it goes with everything. You put this on, you are good to go!\",\n","   'Hell I have never been good to go, so, what the hell why not try some jewellery huh?',\n","   'Oh man! You are so wearing that bracelet!',\n","   'I so am',\n","   'You have any idea what this will do to your sex life?'],\n","  'context_speakers': ['JOEY', 'CHANDLER', 'JOEY', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I am telling you, when she leaned back, I could see her brain',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey',\n","   'So how was Joan?',\n","   'Broke up with her',\n","   'Why',\n","   \"Don't tell me, because of the big nostril thing?\"],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'CHANDLER', 'MONICA', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Dee. The sarcastic sister from What's Happening.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Could you want her more?', 'Who?'],\n","  'context_speakers': ['CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, um, I don't care!\",\n","  'speaker': 'CHANDLER',\n","  'context': ['How do I look?'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I was making a coconut phone with a professor.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Where the hell have you been?'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Well, y'know I'm 29. I mean who needs a savings account.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Well, you're amazing.\",\n","   'Oh no-no-no.',\n","   \"A TV as if it appears from nowhere! That's the dream! Man, how did you afford this stuff?\"],\n","  'context_speakers': ['JOEY', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'It was an accident. Not like I was across the street with a telescope and a box of donuts.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Still upset because I saw her boobies.',\n","   'Well what? What were you doing seeing her boobies?'],\n","  'context_speakers': ['CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I think this is probably why Damone smokes in his cell alone.',\n","  'speaker': 'CHANDLER',\n","  'context': ['I just want to go back to my cell, cause in my cell I can smoke.',\n","   'Smoke away!'],\n","  'context_speakers': ['JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'You know some might find it amusing, I myself find it regular.',\n","  'speaker': 'CHANDLER',\n","  'context': ['What an interesting approach to guitar instruction.'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"I'm sorry man there is never a good time to have to .. Stop catching on fire.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"When I had insurance I could get hit by a bus or catch on fire, it wouldn't matter. Now I got to be careful?\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Hes not right for the part, and if I suggest him my bosses are going to think I am an idiot. And thatÕs something they should learn on their own.',\n","  'speaker': 'CHANDLER',\n","  'context': ['What am I gonna do now?', 'Just pass the tape along.'],\n","  'context_speakers': ['CHANDLER', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Gee, if only she were one and had no idea what a birthday was.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"You can't go away this weekend, its Emma's birthday, we are having a party.\",\n","   \"Well can't you just have the party when we get back?\",\n","   \"No, it won't be her real birthday.\"],\n","  'context_speakers': ['RACHEL', 'MONICA', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Yeah. Yeah, that's what you need a good .. pill.\",\n","  'speaker': 'PHOEBE',\n","  'context': ['Yeah.',\n","   \"I go see my doctor tomorrow, I'll ask her about this. Maybe she can give me a pill or something.\"],\n","  'context_speakers': ['PHOEBE', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Sounds like she has got the whole package!',\n","  'speaker': 'MONICA',\n","  'context': ['How was your date with Ginger.',\n","   'Great. It was great. She is , uh, she is great, great looking, great personality, she is greatness'],\n","  'context_speakers': ['MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Maybe he was nervous.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Whoa ostralopificus isn't supposed to be in that display. No, no, Homohabilus was erect, Ostralopificus was never fully erect.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Nooo! No, that's not true. No, best time ever! Yeah, you rocked my world!\",\n","  'speaker': 'MONICA',\n","  'context': ['Do you know what just happened?',\n","   'Yeah. We-we had sex and then we fell asleep.',\n","   'No. We were in the middle of sex .. and'],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Well thank God your livelihood doesn't depend on it.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey, I need you to set me up for a joke. Later when Monica is around I need you to ask me about firetrucks',\n","   \"I don't know Chandler, I am not so good with remembering lines.\"],\n","  'context_speakers': ['CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Yeah either that or gloria estefan was right eventually the rhythm is going to get you',\n","  'speaker': 'CHANDLER',\n","  'context': ['Oh Mommy, Oh Daddy, I am a big ol Baddy.',\n","   'I guess he must have gotten the part in the play.'],\n","  'context_speakers': ['JOEY', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Ah you see perfect might be a problem. Had you said co-dependent or self-destructive.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Because I met somebody who would be perfect for you.'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I really, really enjoyed it. Very exotic.',\n","  'speaker': 'PHOEBE',\n","  'context': [\"Ha-ha, it's not my baby, ha-ha-ha!\", 'Very good,  so good.'],\n","  'context_speakers': ['PHOEBE', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'If only there was something in your head to control the things you say.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Remember what happened the last time I did an interview for them? I said I write a lot of my own lines, then the writers got mad and made my character fall down the elevator shaft. So who knows what I might say this time.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Yes, we all know how cruel a parent can be, about the flatness of a child's pillow.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"I just don't wanna give them any more ammunition than they already have.\"],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Wow you look just like your son, Mrs. Tribbiani',\n","  'speaker': 'CHANDLER',\n","  'context': ['[Laughs]'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Maybe it\\'s a contest, you know? Like \"Collect all five\".',\n","  'speaker': 'CHANDLER',\n","  'context': ['I know, I opened it up and it  and there it was just floathing in there, like this tiny little hitchhiker.'],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'But younger than some buildings!',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I know Richard's really nice and everything, it's just that we don't know him really well you know, plus he is old  ...  er than some people.\"],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Nope, that patch is no good.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Okay thatÕs like the least fun game ever.',\n","   'Well I am really sick of your smoking. So I brought something that is going to help you quit.'],\n","  'context_speakers': ['CHANDLER', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"No its not big, not at all. Kind of the same lines as, I don't know, having a third nipple!\",\n","  'speaker': 'MONICA',\n","  'context': ['You are dead meat', \"I didn't know it was a big secret.\"],\n","  'context_speakers': ['MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Was that place the sun?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hold on, there is something different.',\n","   'I went to their tanning place your wife suggested.'],\n","  'context_speakers': ['CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'You building a post office?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey, we know how we always say we need a place for the mail? Well I started building one, but then, I decided to take it to the next step.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Did I wake you?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Are you kidding me? Joey Joey Joey Joey!'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'The British are coming?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey you guys, guess what?'],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Yeah okay, I was just talking to the guys. Just look at them I mean, is it okay if they come visit?',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Ahh! Aren't you dressed yet?\",\n","   'Am I naked again?!',\n","   \"We're supposed to meet my parents in 15 minutes.\"],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'You sure he is going to be able to crack that code?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Well, despite the yummy bagels and palpable tension, I have got pants that need to be altered.'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Uh, okay. The fifth dentist caved and not they're all recommending Trident?\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Who?',\n","   \"Dee, the sarcastic sister from What's Happening!!\",\n","   'Guys, guess what guess what guess what.'],\n","  'context_speakers': ['ROSS', 'CHANDLER', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Are you still enjoying your nap?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Okay?', \"Hi Emma, it's the year 2020.\"],\n","  'context_speakers': ['ROSS', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Okay, are you aware unagi  is an eel?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Chandler. I sensed it was you.',\n","   'What?!',\n","   \"Unagi. I'm always aware.\"],\n","  'context_speakers': ['ROSS', 'CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Joey ate my last stick of gum, so I killed him. Do you think that was wrong?',\n","  'speaker': 'CHANDLER',\n","  'context': ['How old is that sock!. If I had a sock on my hand for thirty years, it would be talking too.',\n","   \"Okay, I think its time to change somebody's nicotine patch.\"],\n","  'context_speakers': ['CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"What-wh-what so we'll just stay married forever?\",\n","  'speaker': 'RACHEL',\n","  'context': [\"Wow! This is so amazing. I uh, I really thought I'd have to talk you into this more.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh I am sorry, do you need a break?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Oh wait stop! Stop! Stop!'],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Was it not pain-free?',\n","  'speaker': 'PHOEBE',\n","  'context': ['Did that!',\n","   \"Grasp one of the linen strips by its  'easy grab tab' and pull it off in one quick pain free motion.\",\n","   'Okay.  Ow!!!!! Ow-oh-oh!'],\n","  'context_speakers': ['MONICA', 'PHOEBE', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Is it me or the greetings gone downhill around here?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey', 'You son of a bitch!'],\n","  'context_speakers': ['CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Huh! Whats dufus speled backwards?',\n","  'speaker': 'MONICA',\n","  'context': ['Okay done',\n","   'Whats pleh?',\n","   'ThatÕs Help spelled backwards so that helicopters can read it from the air.'],\n","  'context_speakers': ['JOEY', 'MONICA', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'For what? For letting you throw me out of your office?',\n","  'speaker': 'ROSS',\n","  'context': ['Hi. Look um, about what happened earlier...',\n","   'No, hey, well, I-I completely understand. You were, you were stressed.',\n","   'I was gonna give you a chance to apologise to me.'],\n","  'context_speakers': ['RACHEL', 'ROSS', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Were you so late because you were burying this woman?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Ok it was just me and her at the back of the train and I sat at the door so she had to pass by me if she like to switch cars. She was totally at my mercy.'],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Did you tell her about the magical night that starts with the flush of every toilet?',\n","  'speaker': 'CHANDLER',\n","  'context': ['So Ross, how was the date the other night?'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Uh, so how many cameras are actually on you.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"Where's Monica?\",\n","   'Over here Dad',\n","   'How to you zoom out? There she is',\n","   'Some girl ate Monica',\n","   'Shut up, the camera adds 10 pounds.'],\n","  'context_speakers': ['PERSON1', 'MONICA', 'PERSON2', 'JOEY', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'You said you spent all that money because those boots go with skirts, dresses and pants. What do you plan on wearing? A Chicken suit?',\n","  'speaker': 'CHANDLER',\n","  'context': [\"And the boots don't really go with it.\"],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Uh yes! Thank you.',\n","  'speaker': 'ROSS',\n","  'context': [\"We're not gonna do that.\",\n","   \"Well, if you like you can feel Rachel's and then feel mine to compare.\",\n","   'Am I interrupting?'],\n","  'context_speakers': ['RACHEL', 'PERSON1', 'PERSON2'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh sure, go with the sissy.',\n","  'speaker': 'JOEY',\n","  'context': [\"Yeah, but Jason's really sensitive.\",\n","   'Well sensitive is important, pick him.',\n","   'Yeah.'],\n","  'context_speakers': ['PHOEBE', 'CHANDLER', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"That's so funny because you look more like Joey Tribbiani man slash woman.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['And this from cry for help department, are you wearing makeup?',\n","   'Yes I am, as of today, I am official Joey Tribbiani actor slash model.'],\n","  'context_speakers': ['CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Ah your own brand of vigilante justice.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Did you call the cops?', 'No, we took her to lunch'],\n","  'context_speakers': ['JOEY', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"A tshirt that says - I don't belong here.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Hey guys, what should I wear to a Nick's game?\"],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Oh, you say someone's name enough, they turn around.\",\n","  'speaker': 'PHOEBE',\n","  'context': ['David who?',\n","   'David the scientist guy, David that I was in love with, David who went to Russia and broke my heart David!',\n","   'Oh my God!'],\n","  'context_speakers': ['MONICA', 'PHOEBE', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"And sometimes I would want you to steal third, and I'll go like this.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"And sometimes I may be away longer, like this. But I'll still always come back, like this.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Actually I yesterday I was smoking again. Today I am smoking still.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Boy do I feel bad',\n","   'very bad.',\n","   'Chandler what are you doing?',\n","   'Chandler?'],\n","  'context_speakers': ['PHOEBE', 'MONICA', 'PHOEBE', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Me? I am gonna die alone!',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey Chandler? Monica just broke my sea-shell lamp.'],\n","  'context_speakers': ['RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Joey's got a really bad hernia but its nothing that a laser eye surgery wont fix\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Thank you, but it would take me forever to pay  you that money back and I don't want that hanging on my head. Besides, soon as my insurance kicks in I' ll all the free operations I want. T'll probably start with that laser eye surgery.\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Like I have nothing better to do!',\n","  'speaker': 'MONICA',\n","  'context': ['Richard!',\n","   \"Actually, I'm not here to complement the chef.\",\n","   \"Ohh ..  Oh, that's okay I hate when people come back to complement the chef.\"],\n","  'context_speakers': ['MONICA', 'PERSON', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh no, now its not gonna make any sense',\n","  'speaker': 'CHANDLER',\n","  'context': ['Oh Oh car car.',\n","   'Its Ross on one of his drives',\n","   'What is he doing here?',\n","   'He is saving your butt. Unless ofcourse I am stepping on some toes here, in which case I can just mosey on. I have plenty of people to help in the interstate.'],\n","  'context_speakers': ['RACHEL', 'PHOEBE', 'RACHEL', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I am sorry, we donÕt have your sheep.',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey'],\n","  'context_speakers': ['RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"I am sorry, it was a one time thing. I was very drunk and it was someone else's subconscious\",\n","  'speaker': 'CHANDLER',\n","  'context': ['You and I were, uh',\n","   'doing it on this table',\n","   'You okay there?',\n","   \"I can't belive you two had sex in her dream.\"],\n","  'context_speakers': ['RACHEL', 'PHOEBE', 'CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"Stating the obvious, but thank you. And it's not weird is it.\",\n","  'speaker': 'PHOEBE',\n","  'context': [\"Yes. Great, let's begin.  How's the pressure?\",\n","   'Nice! Wow Phoebe you are good!'],\n","  'context_speakers': ['PHOEBE', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'I just have to do one thing, really quickly, its not a big thing  ...  GET UP||',\n","  'speaker': 'CHANDLER',\n","  'context': ['Get up',\n","   'What?',\n","   \"You're in my seat\",\n","   'How is it your seat?',\n","   'Cause I was sitting there',\n","   'But then you left',\n","   'But its not like I went to Spain, I went to the bathroom , you knew I was coming back.',\n","   'Whats the big deal?',\n","   \"The big deal is that I was sitting there last, so.. that's my seat\",\n","   'Actually, the last place you were sitting was in there',\n","   \"You guys, you know what? it doesn't matter. Because you both have to go get dressed before the big vein in my head pops!\"],\n","  'context_speakers': ['CHANDLER',\n","   'JOEY',\n","   'CHANDLER',\n","   'JOEY',\n","   'CHANDLER',\n","   'JOEY',\n","   'CHANDLER',\n","   'JOEY',\n","   'CHANDLER',\n","   'JOEY',\n","   'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': 'You have to stop the Q-Tip when there is resistance',\n","  'speaker': 'CHANDLER',\n","  'context': [\"My family's been going to him forever. He did my first suit when I was 15. No wait 16. Excuse me, 15 ...  Alright when was 1990?\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': True},\n"," {'utterance': \"You'd better act surprised.\",\n","  'speaker': 'JOEY',\n","  'context': ['About what?',\n","   \"He's planning your birthday party.\",\n","   'Oh my God! I love him!'],\n","  'context_speakers': ['MONICA', 'JOEY', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What was that?!',\n","  'speaker': 'RACHEL',\n","  'context': ['Rachel is one of my closest friends. Although, being the only one who knows anything about this does makes me feel special. Okay!',\n","   'Okay, stop-stop! Phoebe?!',\n","   'Hey Rach!'],\n","  'context_speakers': ['PHOEBE', 'RACHEL', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Seriously, it's moving!\",\n","  'speaker': 'RACHEL',\n","  'context': [\"Hey you guys I'm gonna go out and take a walk.\",\n","   'Phoebe, why is your bag moving?',\n","   \"Oh, it's not!\"],\n","  'context_speakers': ['PHOEBE', 'ROSS', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'I studied for a year in Barcelona.',\n","  'speaker': 'PERSON',\n","  'context': [\"It's from France .. In Europe .. Western Europe. Y'know umm, a few years ago I actually was backpacking across Western Europe.\",\n","   'Really?',\n","   'Uh-hmm,Wait! It gets better. Um, yeah I was in Barcelona.'],\n","  'context_speakers': ['ROSS', 'PERSON', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I mean if you hadn't initiated it I-I-I know I would've.\",\n","  'speaker': 'ROSS',\n","  'context': [\"Yeah. Listen about that, the whole uh, who came on to who thing really doesn't matter. I mean, I think it would've happened either way.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I really don't wanna sit with Allen Iverson over there .\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Wish I could switch with someone.'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Paper cut!',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I can't figure this out! It's so hard! Should I get her a  Tiffany cut or a Princess cut or a,ah-ah!\"],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Well, we're trying to find someone to perform our wedding and they're all either boring or annoying or y'know, can't stop staring at the ladies.\",\n","  'speaker': 'MONICA',\n","  'context': ['Well then we still have a problem.', 'Yeah!', 'With what?'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Thank you.',\n","  'speaker': 'MONICA',\n","  'context': [\"Yeah??? Oh... Okay, yes, we'll be right, we'll be right down.\"],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Ow, ow, ow, ow, leg cramp, leg cramp, leg cramp.',\n","  'speaker': 'PERSON',\n","  'context': [\"No-no-no. I don't have a son named Jordie. We all agreed, my son's name is Jamie.\",\n","   \"Well, Jamie was the name of Susan's first girlfriend, so we went back to Jordie.\",\n","   'What? Whoa, whoa whoa whoa, what do you mean,'],\n","  'context_speakers': ['ROSS', 'PERSON', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"But I'm the host!\",\n","  'speaker': 'JOEY',\n","  'context': ['This will be your phone.',\n","   \"That's great. But uh, I'm not really expecting a lot of calls.\",\n","   'No you answer it and take pledges.'],\n","  'context_speakers': ['PERSON', 'JOEY', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'We- we had our first fight this morning.',\n","  'speaker': 'ROSS',\n","  'context': ['Do you always have to bring him here?',\n","   \"I didn't wanna leave him alone. Alright?\"],\n","  'context_speakers': ['MONICA', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I'm the ruptured spleen.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Yeah, come on, eat, whatever you want. Dr. Roger got beeped again.',\n","   'Yeah I know, guess who beeped him?',\n","   'What?!'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'That oughta do it.',\n","  'speaker': 'PHOEBE',\n","  'context': [\"Where's Chandler?\",\n","   'Oh, he needed some time to grieve.',\n","   \"I'm free! I'm free!\"],\n","  'context_speakers': ['RACHEL', 'PHOEBE', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Yeah! Look!',\n","  'speaker': 'JOEY',\n","  'context': [\"Oh but look! That's gonna leave a stain!\",\n","   \"Rach! Hey! It's fine! You're at Joey's!\",\n","   'Really?'],\n","  'context_speakers': ['RACHEL', 'JOEY', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'And it came out, \"Squatternut buash?\"',\n","  'speaker': 'JOEY',\n","  'context': ['Hey, I wasn\\'t the only one who looked like an idiot. All right? Remember when Ross tried to say, \"Butternut squash?\"'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Yeah, she's fired.\",\n","  'speaker': 'ROSS',\n","  'context': ['Was it my work? Were they mad? Was it Jack? Did he yell?!',\n","   'J-j-just relax, nobody yelled. Jack just was calling to make sure that you were feeling better.',\n","   'Thank God.'],\n","  'context_speakers': ['PHOEBE', 'ROSS', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"It's not funny!\",\n","  'speaker': 'ROSS',\n","  'context': [\"That only is not funny, it's physically impossible! Okay? Depending    on the species I'd have to have a six foot long...\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'I did not mean to touch that,I mean',\n","  'speaker': 'RACHEL',\n","  'context': [\"Okay, well then how about a handshake? Oh God I'm sorry! Oh God, I'm sorry!\"],\n","  'context_speakers': ['RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"It's that whole sensitive thing.\",\n","  'speaker': 'JOEY',\n","  'context': [\"It's gonna be worth it. It's a known fact that women love babies, all right? Women love guys who love babies.\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Hey, does somebody wanna hand me one of those tiles.',\n","  'speaker': 'JOEY',\n","  'context': ['Have you seen Joey?',\n","   \"What's the matter?\",\n","   \"Oh, just this!  Y'know what it's my fault really, because the couch is usually where we\"],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"See Joe that's why your parents told you not to jump on your bed!\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Whoa!'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Okay, Janice. Janice. You gotta give me Janice. That wasn't about being picky.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Okay, Mr. Heckles, we'll try to keep it down.\",\n","   \"Thank you. I'm going to rejoin my dinner party.\",\n","   'All right, bye-bye.'],\n","  'context_speakers': ['MONICA', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Well, umm, my Mom's friend, Phoebe, is actually my birth Mom.\",\n","  'speaker': 'PHOEBE',\n","  'context': [\"Oh, I'll go out and get you some.\",\n","   'Really?!',\n","   \"Nope! Because I'm not your boyfriend. Hey Pheebs, how did it go?\"],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'See, there you go, the cleansing works!',\n","  'speaker': 'PHOEBE',\n","  'context': ['So, um, will you bring the truck?',\n","   \"I'll even let you ring the bell.\",\n","   'Oh, my god.'],\n","  'context_speakers': ['MONICA', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'No!! You have been screwing us all day!',\n","  'speaker': 'MONICA',\n","  'context': [\"Oh you've got to be kidding me.\",\n","   'What?',\n","   'As a wedding gift to you, the hotel would like to give you the honeymoon suite.'],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Oh, hey, look, I found coffee! Okay, let's skedaddle.\",\n","  'speaker': 'PHOEBE',\n","  'context': ['Okay, let me just get a cup of coffee.',\n","   \"Oh Rachel, I know the best coffee house and it's sooo close.\",\n","   'Closer than here?'],\n","  'context_speakers': ['RACHEL', 'MONICA', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Wow! Imagine what our kids would look like!',\n","  'speaker': 'MONICA',\n","  'context': ['You guys make a', 'Yeah, we look great together.', 'Okay.'],\n","  'context_speakers': ['PHOEBE', 'JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Okay?',\n","  'speaker': 'PHOEBE',\n","  'context': [\"Listen, I'm sorry about that whole thing with Roger. It really wasn't right, and I, and I want to make it up to you, so umm, I brought you something that I think you'll really enjoy. Now, this is just a loan.\"],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'A lonely grey couch. \" O look \" cried Ned. And then the kingdom was his. The End.',\n","  'speaker': 'CHANDLER',\n","  'context': ['It was summer. And it was hot. Rachel was there.'],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Umm, would you please excuse me for a moment?',\n","  'speaker': 'ROSS',\n","  'context': [\"I need to talk to you, it's pretty urgent. It's about Monica and Chandler.\",\n","   'Oh my God! Of course, of course.'],\n","  'context_speakers': ['PHOEBE', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Nobody sees the king!',\n","  'speaker': 'PERSON',\n","  'context': [\"Yeah, I talked to you on the phone, I'm the lady that got stuck with the racecar bed.\",\n","   \"Look, it's like I told you, there's nothing I can do. You signed for it, Monica Velula Geller.\",\n","   'All right, Jester man, look we wanna see the king.'],\n","  'context_speakers': ['MONICA', 'PERSON', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I know but it's so hard! Nothing rhymes with your stupid name!\",\n","  'speaker': 'PHOEBE',\n","  'context': [\"Pheebs, that's great!\",\n","   'Oh, yay!',\n","   \"But y'know umm, Rachel doesn't rhyme with draddle.\"],\n","  'context_speakers': ['RACHEL', 'PHOEBE', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Take a couple deep breaths.',\n","  'speaker': 'JOEY',\n","  'context': ['Okay, I can do this. I can tell her how I feel. Just uh, just stand up straight.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Ah, oh, hang on a second. Come in, come in. All right, so think about it, and call me back.',\n","  'speaker': 'PERSON',\n","  'context': [\"It has umm, three beautiful bedrooms and ah, no baths. But y'know, the ocean is right there.\",\n","   'Knock, knock, knock.'],\n","  'context_speakers': ['PERSON', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Hey, my father's house does that.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['And! And!', 'Ah!'],\n","  'context_speakers': ['PHOEBE', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Stick to the list.',\n","  'speaker': 'ROSS',\n","  'context': ['Ahh!',\n","   \"It wasn't on your list, but hopefully you'll think it's really fun.\",\n","   'A scooter!'],\n","  'context_speakers': ['RACHEL', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"This-this, no, oh no, no-no-no, this is not,that's-that's not what it is.\",\n","  'speaker': 'RACHEL',\n","  'context': [\"Ohh! It's so nice to meet you.  Hello. Hello.\",\n","   'Well, Joshua, that 500 was for groceries.',\n","   'What?'],\n","  'context_speakers': ['RACHEL', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Uh, listen, I need that broiling pan that Joey borrowed the other day.',\n","  'speaker': 'MONICA',\n","  'context': ['How many times have I told you guys, you',\n","   'Hi Chandler.',\n","   'Hey!'],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Thank you.',\n","  'speaker': 'ROSS',\n","  'context': [\"I have a question. Umm, I used your pen to draw on my friend's face. A beard and a moustache.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Oh! Oop!',\n","  'speaker': 'PHOEBE',\n","  'context': [\"It's the big night! We wanted to wish you good luck!\",\n","   'Yeah, yeah you have the ring?',\n","   'Yeah, right here in my pocket. Pheebs?'],\n","  'context_speakers': ['JOEY', 'ROSS', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"So to sum up, I'm Professor Geller.\",\n","  'speaker': 'ROSS',\n","  'context': ['Whoa Let me start by uh, by introducing myself, I am Professor Geller.'],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Nope. Op, wait! There he goes.',\n","  'speaker': 'JOEY',\n","  'context': ['And you can see me! Same as yesterday, same as the day before.',\n","   'Is he doing his shark attack bit yet?'],\n","  'context_speakers': ['JOEY', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'I was thkinking about maybe going into the Robot.',\n","  'speaker': 'ROSS',\n","  'context': [\"Hey, you haven't been practising the routine have you?\", 'No!'],\n","  'context_speakers': ['MONICA', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"It's not a real game! I made it up!\",\n","  'speaker': 'CHANDLER',\n","  'context': ['And now you want that money back.',\n","   'Exactly.',\n","   'Chandler, what kind of an idiot do you take me for?'],\n","  'context_speakers': ['ROSS', 'CHANDLER', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Oh just like give a minute to wake up for this  ...  Ha ha! You lost the ring. You the worst best man ever!',\n","  'speaker': 'CHANDLER',\n","  'context': ['The stripper stole the ring! The stripper stole the ring! Chandler get up get up, the stripper stole the ring.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'No, I know, but I,he just went on.',\n","  'speaker': 'RACHEL',\n","  'context': ['May I see your boarding pass?',\n","   \"Oh, no, no, I don't have one. I just need to talk to my friend.\",\n","   \"Oh, oooh. I'm sorry. You are not allowed on the jetway unless you have a boarding pass.\"],\n","  'context_speakers': ['RACHEL', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Pheebs?',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I can't take it anymore. So you win, okay? Here!\"],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Anyway, I should go, one of the lifeguards was just about to dismantle a nuclear device.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I know, it is isn't it?\",\n","   'No, I mean it, this feels really good. Is it a hundred percent cotton?',\n","   'Yeah! And I got it on sale, too.'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I'm supposed to teach a graduate seminar there in ten minutes.\",\n","  'speaker': 'ROSS',\n","  'context': [\"Okay, that-that's all for today. Oh, uh does anyone know where the Freeman building is?\",\n","   \"Yeah, it's the new building on Avenue A.\"],\n","  'context_speakers': ['ROSS', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I'll catch you guys later.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Hi, excuse me, is Rachel Green here? I'm supposed to meet her for lunch.\",\n","   \"Oh she doesn't come here anymore. You can find her up on 10th.\",\n","   'Okay great.'],\n","  'context_speakers': ['CHANDLER', 'PERSON', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Chloe, switch with me, there's some guys here that got a crush on you.\",\n","  'speaker': 'PERSON',\n","  'context': ['Can I help you?',\n","   'Uh-oh.',\n","   \"Uh, y'know what, we're having second thoughts about our copying needs. And we'll need a little more time to think about it.\"],\n","  'context_speakers': ['PERSON', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Look, please, please don't go anywhere, okay?\",\n","  'speaker': 'PERSON1',\n","  'context': [\"Dr. Farber, we've got a bit of an emergency here...Jason Costalano is choking on his retainer.\",\n","   \"Oh God... I'll be right there, Bernice.\"],\n","  'context_speakers': ['PERSON2', 'PERSON1'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Well, I like it. Here you go.',\n","  'speaker': 'JOEY',\n","  'context': [\"We've got uh, scarves, tulip post cards...\",\n","   \"Check this out? Huh?  Yeah. That's the stuff. What do you think?\",\n","   'Well, I don\\'t have to buy that, \"I\\'m with stupid\" T-shirt anymore.'],\n","  'context_speakers': ['PERSON', 'JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Oh-oh, that's a risky little game!\",\n","  'speaker': 'RACHEL',\n","  'context': ['Oh!', 'Oh God', 'Now you know how you really feel about it.'],\n","  'context_speakers': ['RACHEL', 'MONICA', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'His legs flail about as if independent from his body',\n","  'speaker': 'CHANDLER',\n","  'context': ['According to Chandler, what phenomenon scares the bejesus out of him?',\n","   'Michael flatly lord of the dance',\n","   'That is correct!',\n","   'The Irish gig guy?'],\n","  'context_speakers': ['ROSS', 'MONICA', 'ROSS', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I'm never gonna get to meet her anyway.\",\n","  'speaker': 'JOEY',\n","  'context': [\"But I can see through your sheet. Yeah, yeah, that's her. But y'know what? Doesn't matter,\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Twenty-one hours, you're a hero.\",\n","  'speaker': 'PERSON',\n","  'context': [\"Y'know what I think it is? I think you've made such a nice home for her over the last nine months that she just doesn't want to leave.\",\n","   'Oh. Look at you making up crap for me. Oh God!'],\n","  'context_speakers': ['ROSS', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Okay?',\n","  'speaker': 'PHOEBE',\n","  'context': [\"Okay, here's what we're gonna do. Okay, I'm gonna break into this mini-van and put it in neutral. You guys push it forward so Ross can drive out of his spot.\"],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I'd love to show ya, but I just tucked her in.\",\n","  'speaker': 'JOEY',\n","  'context': ['Hi Joey.',\n","   \"Hey! How you doin'?\",\n","   'He has the most amazing Porsche under there!'],\n","  'context_speakers': ['PERSON', 'JOEY', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Ross umm, there's something that I've got to tell you, there's-there's someone else.\",\n","  'speaker': 'PERSON',\n","  'context': ['It was dreadful. I felt terrible about how I acted when you said those wonderful things.',\n","   \"No, no, that-that, that's all right. Umm, I'm just glad you called.\"],\n","  'context_speakers': ['PERSON', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Ok!',\n","  'speaker': 'JOEY',\n","  'context': ['You know? Forget it!',\n","   'No-no-no-no, no! Who, who were you talking about?',\n","   \"No, I-I-I-I don't, I actually don't know\"],\n","  'context_speakers': ['RACHEL', 'JOEY', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'And I just won a million dollars!',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I don't want to be single. I just, I just want to be married again.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I'm the supply manager.\",\n","  'speaker': 'PERSON',\n","  'context': ['Okay, I can do that! Oh, by the way,',\n","   \"I Why don't we do a trial run.\",\n","   'Oh okay. Umm, all right.  Hi, this is Phoebe from Empire Office Supplies, can I speak to your supply manager please?'],\n","  'context_speakers': ['PHOEBE', 'PERSON', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"And then and then you clicked it again, she's dressed. She is a business woman, she is walking down the street and oh oh oh she's naked.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Did you have fun at the bachelor's party last night?\",\n","   'Oh yeah, look what I got, she is fully dressed right. And then you click it and oh oh she naked.'],\n","  'context_speakers': ['RACHEL', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Just a sec, give me a minute to wake up for this,Ah-ha-ha!!',\n","  'speaker': 'CHANDLER',\n","  'context': ['What?', 'The ring is gone!', 'Ugh.'],\n","  'context_speakers': ['CHANDLER', 'JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'We got a box.',\n","  'speaker': 'JOEY',\n","  'context': [\"When that guy was robbing us and I was locked in the entertainment unit for like six hours, do you know what I was doing there in all that time? I was thinking about how I let you down. Yeah. But if I had known what kind of friend you would turn out to be, I wouldn't have worried about it so much. See you around!\",\n","   'Wait! Just wait one second. There has to be something that I can do, something. If we still had that entertainment unit, I would get in there for six hours and think about  how I let you down.'],\n","  'context_speakers': ['JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Okay, you guys, just relax.  I doooo.',\n","  'speaker': 'RACHEL',\n","  'context': ['No, yeah! The groom cannot see the bride!',\n","   \"I'm not gonna marry Chandler!\",\n","   'Not after this!'],\n","  'context_speakers': ['PHOEBE', 'RACHEL', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'The end.',\n","  'speaker': 'PHOEBE',\n","  'context': ['Oh, but he was my pig man...how did I not see this?',\n","   \"Oh! I know! It's because... he's gorgeous, and he's charming, and when he looks at you...\",\n","   'Ok, Ok, Pheebs...'],\n","  'context_speakers': ['RACHEL', 'PHOEBE', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Oh my God!! Ohh, that is it! I'm leaving! You are just a horrible person!\",\n","  'speaker': 'RACHEL',\n","  'context': [\"Oh, I must've said that after you left.\",\n","   'Said what? Exactly.',\n","   'That you enjoyed the occasional drink .. ing binge.'],\n","  'context_speakers': ['PERSON', 'RACHEL', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"What's going on?!  That's what's going on!!\",\n","  'speaker': 'ROSS',\n","  'context': [\"Okay, okay look, I know I'm being Mr. Inappropriate today, but it's just so tough, I mean see you walking around and I just wanna touch you and hold you, come on no one's around, just, just kiss me.\",\n","   \"All right that's, that's it!! Get off her!\",\n","   'What is going on?'],\n","  'context_speakers': ['PERSON', 'ROSS', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Yes it is.',\n","  'speaker': 'PERSON',\n","  'context': ['Fire alarm? Oh! Hi, officer,fireman, can-can I help you?',\n","   \"We found your fire alarm in the trash chute. That's not mine.\"],\n","  'context_speakers': ['PHOEBE', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Y'know what I mean, like dubbies?\",\n","  'speaker': 'RACHEL',\n","  'context': [\"Y'know what, actually ..   Okay, okay, okay, what's so funny over here?\",\n","   \"I thought you didn't smoke.\",\n","   \"Oh, I thought you guys meant marijuana cigarettes, y'know?\"],\n","  'context_speakers': ['RACHEL', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Ah well, this is the living room.',\n","  'speaker': 'PERSON1',\n","  'context': [\"Really?! Well, it's just like everyone else's apartment. It's got rooms, walls, and ceilings.\",\n","   'Well, I just wanted to see where you lived. Now, give me the tour.',\n","   'Oh my God! Oh my God!'],\n","  'context_speakers': ['PERSON1', 'PERSON2', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What a baby.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I mean, you two are-are my… I mean, I'm lucky to have just one good\",\n","   'Thanks man.',\n","   'I gotta go check something over here.'],\n","  'context_speakers': ['ROSS', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What did you want to show us? Because all I can see is this bitchin van',\n","  'speaker': 'ROSS',\n","  'context': ['Come on, no peeking!',\n","   \"Our eyes are closed and we're about to cross the street. Very good.\",\n","   'Okayyyyy, open up!'],\n","  'context_speakers': ['PHOEBE', 'CHANDLER', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Oh hey, great, you're up. Rachel, this is my sister Krista. Krista, this is Rachel.\",\n","  'speaker': 'PERSON',\n","  'context': [\"I'm really glad Monica asked us out.\",\n","   \"I'd love to ask you in, but uh, my sister's visiting and I think she's asleep on the couch.\",\n","   \"You're sister? You're sister's asleep on the couch? Ohhh! I saw\"],\n","  'context_speakers': ['RACHEL', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'DANGER',\n","  'speaker': 'ROSS',\n","  'context': [\"I don't like sitting up here! I'm just gonna over\",\n","   'No Rachel! They got here first! Why is Ross doing that?'],\n","  'context_speakers': ['RACHEL', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'No just, nobody press on my stomach.',\n","  'speaker': 'JOEY',\n","  'context': [\"Well Joey, we're all .. we're all very proud of you.\",\n","   'Yes, I believe we can expect a call from the President any moment now.',\n","   'Is there anything we can do for you?'],\n","  'context_speakers': ['MONICA', 'CHANDLER', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Hey! It's starting to snow.\",\n","  'speaker': 'ROSS',\n","  'context': [\"No! You let go! How come it's always us left in the field holding the ball?\",\n","   \"I don't know. I guess the other people just don't care enough.\"],\n","  'context_speakers': ['ROSS', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What?!?!',\n","  'speaker': 'MONICA',\n","  'context': ['Well, you-you coulda just turned the cushion over.',\n","   \"Yeah, I would've except I had a big spaghetti stain on the other side.\"],\n","  'context_speakers': ['MONICA', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Uh-huh, well! But umm, still I'm-I'm mad at you.\",\n","  'speaker': 'PHOEBE',\n","  'context': ['I knew it, wow!!',\n","   'Wait-wait-wait, wait! Puppies. Cute or ugly?',\n","   'Ohh, so cute.'],\n","  'context_speakers': ['PERSON', 'PHOEBE', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"No no, you don't have to fly back, really.\",\n","  'speaker': 'MONICA',\n","  'context': [\"Yeah, Ross is great. He's uh, he's in a whole other place. No, he's gone.\"],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I don't think you could ever do that.\",\n","  'speaker': 'PHOEBE',\n","  'context': [\"There is nothing wrong with Jake! Okay? He is all man! I'm thinking even more than you.\",\n","   'Oh yeah, he looked like a real lumberjack in those pink lacys.',\n","   \"I'm just saying that only a man completely secure with his masculinity could walk around in women's underwear!\"],\n","  'context_speakers': ['PHOEBE', 'JOEY', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Nothing. It's just old and dingy, that's all.\",\n","  'speaker': 'JOEY',\n","  'context': ['Oh Monica, listen, I ah, I saw down at the hardware store, they got those designer tiles on sale. If you ever want to redo the bathroom floor.',\n","   \"Why, what's wrong with my bathroom floor?\"],\n","  'context_speakers': ['JOEY', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Yes, I know, but her friend sounds like such a...',\n","  'speaker': 'CHANDLER',\n","  'context': ['Oh, uh, listen, about tonight...',\n","   \"No, no, no, don't you dare bail on me. The only reason she's goin' out with me is because I said I could bring a friend for her friend.\"],\n","  'context_speakers': ['JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'When? Today? Some, not a lot.',\n","  'speaker': 'JOEY',\n","  'context': ['Joe?', 'Yeah!', \"Uh, you've had a lot of sex right?\"],\n","  'context_speakers': ['CHANDLER', 'JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Yeah, a slice of cheesecake and-and a date if you're given'  'em out.\",\n","  'speaker': 'PERSON',\n","  'context': ['You gotta help me out here, you gotta set me up, you gotta get me back in the game.',\n","   \"Well, that shouldn't be a problem. I mean I work in fashion and all I meet are eligible straight men.\",\n","   'Pete, can I get you something else?'],\n","  'context_speakers': ['MONICA', 'RACHEL', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Then we went to this bar and he hooked us up with all these women!',\n","  'speaker': 'PERSON',\n","  'context': ['Hi Tag! Hey, so did you have fun with uh, with Joey last night?',\n","   \"Oh yeah! We went to the Knicks game. Ohh that's nice.\"],\n","  'context_speakers': ['RACHEL', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Oh yeah he has a caretaker his older brother, Ernie. You can't make this stuff up!\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Helo! Anybody in there order a celebrity?',\n","   'What was that?',\n","   'Oh he is just a crazy guy who roams the halls here. He is great with kids though!',\n","   'Is he alright out there by himself?'],\n","  'context_speakers': ['JOEY', 'PERSON', 'CHANDLER', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Oh, no, no. I just meant hypothetically.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I'll tell you what. When we're 40, if neither one of us are married, what do you say you and I get together and have one?\",\n","   \"Why won't I be married when I'm 40?\"],\n","  'context_speakers': ['CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Carol called me to thank me for the lasagna, I asked, she told me.',\n","  'speaker': 'MONICA',\n","  'context': [\"I mean, I couldn't not know, I mean, if, if the doctor knows, and Carol knows, and Susan knows.... And Monica knows…\",\n","   \"Wha, heh, how could you know, I don't even know!\"],\n","  'context_speakers': ['RACHEL', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'It was like months ago.',\n","  'speaker': 'RACHEL',\n","  'context': [\"Oh, that's Ross's.\",\n","   'Oh...  Oh my God. He remembered.',\n","   'Remembered what?'],\n","  'context_speakers': ['CHANDLER', 'RACHEL', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Yeah, I'm fine.\",\n","  'speaker': 'RACHEL',\n","  'context': [\"Hi. Well, look, I was just gonna leave a message, isn't tonight your, your big anniversary dinner?\",\n","   'Yeah. Well, umm.....',\n","   'Rach, are you okay?'],\n","  'context_speakers': ['PERSON', 'RACHEL', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"When they're hungry enough, they'll come in.\",\n","  'speaker': 'JOEY',\n","  'context': ['Yes. Yes, there is, they play on Sundays and Monday nights.',\n","   'Oh shoot! I work Monday nights.',\n","   'Umm, this stuffing is amazing. Do you think we should bring them some?'],\n","  'context_speakers': ['CHANDLER', 'RACHEL', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Well, I believe the piece of furniture was fine until your little breakfast adventure with Angela Delvecchio',\n","  'speaker': 'CHANDLER',\n","  'context': [\"That's the rule.\",\n","   \"What rule? There's no rule, if anything, you owe me a table!\",\n","   \"How'd you get to that?\"],\n","  'context_speakers': ['JOEY', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Yeah. What-what should I do?',\n","  'speaker': 'ROSS',\n","  'context': [\"I mean technically it's-it's not against the rules or anything, but it is frowned upon. Especially by that professor we ran into last night, Judgey von Holierthanthou.\",\n","   \"Well Ross, you be careful now. You don't want to get a reputation as y'know Professor McNailshisstudents.\"],\n","  'context_speakers': ['ROSS', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'So Ross, how umm, how did it go with Bonnie?',\n","  'speaker': 'PHOEBE',\n","  'context': [\"Oh, about five minutes. Right now they're interviewing his opponent. Apparently he trains by going to Iran and pulling the arms off thieves.\",\n","   'Hot dog?',\n","   \"Four, please. I'm really nervous. Thank you.\"],\n","  'context_speakers': ['MONICA', 'ROSS', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Take your shirt off, and let's see what we're dealing with here.  What are you doing?\",\n","  'speaker': 'PERSON',\n","  'context': ['Just your run-of-the-mill third nipple. You can take it off. Just slice that baby right off!'],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Okay, does this look like a garbage can to you?',\n","  'speaker': 'PHOEBE',\n","  'context': [\"Nobody respects the bucket! You wouldn't believe what people put in here! Look!\"],\n","  'context_speakers': ['PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What is the matter with your face?!',\n","  'speaker': 'MONICA',\n","  'context': [\"Let's try a look .. of far off .. wonderment. Okay, we'll-we'll gaze into our future and we'll think about our marriage and the days to come. Chandler!\"],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"So, you're just Bing?\",\n","  'speaker': 'JOEY',\n","  'context': [\"No, no, you're right, it is a ridiculous name!\",\n","   \"It's not that bad.\",\n","   'Yes it is! From now on, I have no first name.'],\n","  'context_speakers': ['CHANDLER', 'JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What?',\n","  'speaker': 'PHOEBE',\n","  'context': ['Oh, I mean she\\'s gonna be at the wedding waiting for him and people will be whispering, \"Oh that poor girl.\" Then she\\'ll have to come back here and live all alone.',\n","   'Oh my God!'],\n","  'context_speakers': ['RACHEL', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'When are you coming home?',\n","  'speaker': 'PERSON',\n","  'context': ['Well, the biggest news is still you dumping Barry at the altar!',\n","   \"Alright. Let's talk reality for a second.\",\n","   'Okay.'],\n","  'context_speakers': ['PERSON', 'PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Wait, wait, Chandler, this is what you're havin' for Thanksgiving dinner?\",\n","  'speaker': 'RACHEL',\n","  'context': ['You owe me 20 bucks.',\n","   'Thank you. Thank you so much!',\n","   \"Chandler, here you go, got your traditional Thanksgiving feast, you got your tomato soup, your grilled cheese fixin's, and your family size bag of Funyuns.\"],\n","  'context_speakers': ['MONICA', 'RACHEL', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I don't know, something girlie.\",\n","  'speaker': 'ROSS',\n","  'context': ['Really, I, no, I think girls night out is a great idea. Okay, okay, bye',\n","   'So what are they doing?'],\n","  'context_speakers': ['ROSS', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I wasn't by myself. I had a doctor, a nurse, and a helper guy. So, did you see who won the game?\",\n","  'speaker': 'PERSON1',\n","  'context': ['Your mom called me. So is this her?',\n","   'No, this is a loaner.',\n","   \"I'm sorry you had to do this by yourself.\"],\n","  'context_speakers': ['PERSON2', 'PERSON1', 'PERSON2'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Weā€™re not peeking?',\n","  'speaker': 'JOEY',\n","  'context': ['Come on!'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What?  Why?',\n","  'speaker': 'MONICA',\n","  'context': ['I do! Want a hint? huh? \"I do\"  \"I do\".',\n","   \"Ok, I'm sensing that this is some kind of word play, because you are pink with barely controlled glee.\",\n","   'David is going to propose to Phoebe.'],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What would you give to a kid if he wanted a ticket to play with?',\n","  'speaker': 'CHANDLER',\n","  'context': ['One ticket to Yemen? Oh no-no-no, no. No, no, no, I just, I just need a pretend ticket.',\n","   \"I'm sorry sir, I don't understand.\"],\n","  'context_speakers': ['CHANDLER', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"So how'd it go?\",\n","  'speaker': 'ROSS',\n","  'context': ['Yeah. He could row like a viking.', 'Hi.'],\n","  'context_speakers': ['ROSS', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Hey, do you know a six-letter word for red?',\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey! Hey! What are you guys doing up?',\n","   'Oh, we wanted to finish the crossword before we went to bed.'],\n","  'context_speakers': ['JOEY', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Am I out of control?',\n","  'speaker': 'MONICA',\n","  'context': [\"Uh, belt's to tight, gotta change the belt. Did I turn the fish? No, cause I made lasagne.\"],\n","  'context_speakers': ['MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Well what are you gonna do when he finds out he wasn't even asked?!\",\n","  'speaker': 'MONICA',\n","  'context': [\"Oh! Right. Umm, maybe that's because I didn't send him an invitation.\",\n","   \"Chandler! He's your father.  he should be at the wedding.\",\n","   \"I don't even know the man. Okay? We're not that close. I haven't seen him in years.\"],\n","  'context_speakers': ['CHANDLER', 'MONICA', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Well, don't  'cha wanna?\",\n","  'speaker': 'MONICA',\n","  'context': [\"It's just y'know, been a couple of hours, and she hasn't called. Not that I even care, y'know.\",\n","   \"Phoebe why don't you just call her? You obviously want to.\",\n","   'You think you know me so well.'],\n","  'context_speakers': ['PHOEBE', 'MONICA', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'You still love me?',\n","  'speaker': 'ROSS',\n","  'context': [\"All right! Ross, do you think it's easy for me to see you with somebody else?\",\n","   \"Y'know, hey! You're the one who ended it, remember?\",\n","   'Yeah, because I was'],\n","  'context_speakers': ['RACHEL', 'ROSS', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'And did he?',\n","  'speaker': 'MONICA',\n","  'context': [\"We couldn't keep our eyes off each other all night and then every once and a while y'know, he'd kinda lean over and stroke my hair and touch my neck.\",\n","   \"Okay, stop it Phoebe, you're getting me all tingly.\",\n","   'All I could think of was y\\'know, \"Is he gonna kiss me? Is he gonna kiss me?\"'],\n","  'context_speakers': ['PHOEBE', 'MONICA', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Me too! There's a game on Tuesday, do you wanna go?\",\n","  'speaker': 'JOEY',\n","  'context': [\"It's Jake.\",\n","   'Joey. Hey Jake, do you like the Knicks?',\n","   'Yeah, big fan.'],\n","  'context_speakers': ['PERSON', 'JOEY', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Do you still wanna call  'em? I wanna call  'em.\",\n","  'speaker': 'MONICA',\n","  'context': ['Yeah! You gotta tell a girl before you tape her. Such a rookie mistake.',\n","   \"Y'know who has a great video camera?\",\n","   'Greg and Jenny?'],\n","  'context_speakers': ['JOEY', 'CHANDLER', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What are you up too?',\n","  'speaker': 'MONICA',\n","  'context': ['Yeah! And also, a little like a French guy. I never noticed that before.'],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Anyway, they want to take us out Saturday night! What do you say?',\n","  'speaker': 'RACHEL',\n","  'context': ['Oh my!', \"They're male nurses.\", 'Not in my head.'],\n","  'context_speakers': ['JOEY', 'RACHEL', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"You're gonna have to tell me how you did that.\",\n","  'speaker': 'RACHEL',\n","  'context': [\"Oh good, you're here. Yeah, and I was worried that it was going to be uncomfortable.\",\n","   \"I know Mona, just hear me out. First of all, I'm so sorry about my father yelling at you, but I heard you totally held your own.\"],\n","  'context_speakers': ['PERSON', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'No',\n","  'speaker': 'PHOEBE',\n","  'context': ['No, no!',\n","   'Give it to me!',\n","   \"You can't have it.\",\n","   'Give it to me!'],\n","  'context_speakers': ['PHOEBE', 'MONICA', 'PHOEBE', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'I do not always have to be,okay, okay.',\n","  'speaker': 'ROSS',\n","  'context': [\"No you do,y'know what, you're not gonna suck me into this.\",\n","   'Oh sure I am, because you always have to be right.'],\n","  'context_speakers': ['ROSS', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Ross-Ross, you have no idea what this means to me!',\n","  'speaker': 'RACHEL',\n","  'context': [\"I would love to live with you Ross.. that's-that's great! Thank you!\",\n","   \"Well, I'm-I'm just glad I could, y'know, help you out.\"],\n","  'context_speakers': ['RACHEL', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"All right, let's do it.\",\n","  'speaker': 'RACHEL',\n","  'context': ['Rachel, I have not missed one question the whole game. I own this game! Look at my hand.',\n","   'Why? Do you have the answers written on there?',\n","   'No! Steady as a rock! Now, are you with me.'],\n","  'context_speakers': ['MONICA', 'RACHEL', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"The big deal is I don't want naked, greasy strangers in my apartment when I want to kick back with a puzzle,beer! Cold beer.\",\n","  'speaker': 'ROSS',\n","  'context': ['Okay, well Ross, what is this really about?',\n","   'Look, this is my home and I want to be able to come and go whenever I want!',\n","   \"Okay, I will find someplace else to do the rest of my appointments. I just don't know what the big deal is!\"],\n","  'context_speakers': ['PHOEBE', 'ROSS', 'PHOEBE'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Y'know you, you just stop being such a wuss and get those off and you come with us and watch naked girls dance around!!\",\n","  'speaker': 'PHOEBE',\n","  'context': [\"No you can't. You don't want to be guys, you'd be all hairy and wouldn't live as long.\"],\n","  'context_speakers': ['CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Yes you can. You're thinking about time, you can't go back in time.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['You know what I knew something like this is gonna happen',\n","   \"What are you doing? Chandler, you can't just go back a phase\"],\n","  'context_speakers': ['CHANDLER', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Okay, but If he asks, I am not going to lie.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"But if you do, make sure it seems like you're there to see him, okay? And you're not, like, doing it as a favor to me.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'No!',\n","  'speaker': 'JOEY',\n","  'context': ['Oh, unbelievable! We had the best time!',\n","   \"Yay!! Oh so, you're not, you're not mad at us anymore?\",\n","   'No!'],\n","  'context_speakers': ['JOEY', 'PHOEBE', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Whoa, maybe I was!',\n","  'speaker': 'JOEY',\n","  'context': [\"Very life-like and funny. Okay. Oh no-no-no, I wasn't waving at you lady.\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"He's right, even if it's to say something complementary.\",\n","  'speaker': 'ROSS',\n","  'context': ['Oh.',\n","   \"Please, right now, no, every time I see him it's like  'Is it on the lose?'  'Is it watching me?'\",\n","   \"We can't tell him, you can't go up to a guy you barely know and talk about his.... stuff.\"],\n","  'context_speakers': ['ROSS', 'PHOEBE', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Okay?',\n","  'speaker': 'JOEY',\n","  'context': [\"I don't know. But hey, I know we can decide. Okay, I'm gonna ask you questions and you answer real quick.\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'My duties?  All right.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"You must've had your hands full\",\n","   'That I did. That I did.',\n","   \"So let's talk a little bit about your duties.\"],\n","  'context_speakers': ['PERSON', 'CHANDLER', 'PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Yeah?',\n","  'speaker': 'PHOEBE',\n","  'context': ['Uh, okay, Pheebs?',\n","   'Yeah?',\n","   \"Umm, y'know how the other day you were talking about how you didn't get to go to London and how you were kinda feeling left out?\"],\n","  'context_speakers': ['JOEY', 'PHOEBE', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Ohhh!!!!',\n","  'speaker': 'RACHEL',\n","  'context': [\"OK, I have a question. Well, actually, it's not so much a question as.. more of a general wondering... ment.\",\n","   'OK.',\n","   \"OK. Here goes. For a while now, I've been wanting to, um....\"],\n","  'context_speakers': ['ROSS', 'RACHEL', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"No, no, no, it's flattery.\",\n","  'speaker': 'ROSS',\n","  'context': [\"Aren't you with that girl over there?\",\n","   \"Well, yeah, kinda. Um, but that's okay, see we have an understanding, um, see we each have this list of five famous people, so I'm allowed to sleep with you.\"],\n","  'context_speakers': ['PERSON', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Yeah!',\n","  'speaker': 'PHOEBE',\n","  'context': ['Oh. Thank you. Ohhh, thank you very much. Oh, thank you for coming.  Uh, just a second!',\n","   'No-no, let me in!',\n","   'Phoebe?'],\n","  'context_speakers': ['MONICA', 'PHOEBE', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Really!?',\n","  'speaker': 'RACHEL',\n","  'context': ['Look honey, Mark is in fashion okay, I like having a friend that I can share this stuff with. You guys would never want to go to a lecture with me.',\n","   'Pa-haa!! I would love to go with you.'],\n","  'context_speakers': ['RACHEL', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': ': Yeah.',\n","  'speaker': 'PERSON',\n","  'context': ['I know, it was amazing! I mean, we totally nailed it, it was beautiful.',\n","   'Thank you, listen, thanks a lot Tribbiani, . Oh my God, look at the time, I gotta catch the bus to the ball.',\n","   'Oh well, okay, good luck.'],\n","  'context_speakers': ['JOEY', 'PERSON', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'No.',\n","  'speaker': 'RACHEL',\n","  'context': [\"Okay, I can't imagine, I can't imagine my life without you. Without, without these arms, and your face, and this heart. Your good heart Rach, and, and....\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Yeah.',\n","  'speaker': 'PERSON',\n","  'context': [\"I mean, your Dad tells me that you get a couple of days off school, and you, you ah, don't have to sell those cookies anymore.\",\n","   'Well, I kinda wanted to sell the cookies. The girl who sells the most wins a trip to Spacecamp, and gets to sit in a',\n","   'Wow, you ah, you really like all this space stuff, huh?'],\n","  'context_speakers': ['ROSS', 'PERSON', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Yeah, she couldn't live without the Chan Love.\",\n","  'speaker': 'CHANDLER',\n","  'context': ['Hey Pheebs!',\n","   'Ohh! You made up!',\n","   \"Yeah, I couldn't be mad at him for too long.\"],\n","  'context_speakers': ['CHANDLER', 'PHOEBE', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Uh-huh!',\n","  'speaker': 'PHOEBE',\n","  'context': [\"Look, Pheebs, I just, I wanna apologize for, for saying that your method was stupid and-and maybe ask you to be my teacher again. And-and I promise, I won't touch a guitar until you say I'm ready. You really think I'm ready?\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Oh yeah. First off all, chasing the Churo guy isn't jogging.\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Oh, y'know the same thing happened to me one time.\",\n","   'When did that happen to you?!',\n","   \"Don't you remember when we were jogging in the park and we saw that really pretty bird and wanted to take a picture,I didn't have my camera!\"],\n","  'context_speakers': ['JOEY', 'CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'That is the coast guard!',\n","  'speaker': 'RACHEL',\n","  'context': [\"Just because he's got a bigger boat he thinks he can take up the whole river. Get out of the way jackass! Who names their boat Coast Guard anyway?\"],\n","  'context_speakers': ['JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Thanks, but I kinda have a problem to deal with here.',\n","  'speaker': 'JOEY',\n","  'context': [\"I didn't know he could get me fired! What am I going to do?\",\n","   \"Well y'know Joey, you're a pretty charming guy.\"],\n","  'context_speakers': ['JOEY', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Because his penis was too big. I am sorry, thatÕs the kind of thing I do.',\n","  'speaker': 'CHANDLER',\n","  'context': ['I am sorry, why did Phoebe and Mike breakup?'],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Look I never should have kissed your girlfriend, but I'm\",\n","  'speaker': 'CHANDLER',\n","  'context': [\"Hey, it's me. I know you can't stand to be in the same room as me, so I just thought I'd try and apologize over the phone.\",\n","   'Hello.'],\n","  'context_speakers': ['CHANDLER', 'JOEY'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Oh no donÕt go. I have scared you. I have said too much. I am hopeless and awkward and desperate for love.',\n","  'speaker': 'CHANDLER',\n","  'context': [\"I am not divorced yet. Chandler, you know you just invited me over here for pasta and all of a sudden you are talking about moving in together. I wasn't even that hungry. Its getting really late and I should just umm.\"],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"I-I'm really sorry. It's just that I\",\n","  'speaker': 'ROSS',\n","  'context': [\"Hi umm, listen come here, come in. I'm so-so sorry about yesterday.\"],\n","  'context_speakers': ['ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Some guy bought it. I'm sorry. I tired to stop it but they put me in jail!\",\n","  'speaker': 'PHOEBE',\n","  'context': [\"Oh, that's uh, that's pretty nice but I'm gonna go with the one I picked first.\",\n","   \"Oh my God Chandler, the one you picked is gone. It's over!\",\n","   'What?'],\n","  'context_speakers': ['CHANDLER', 'PHOEBE', 'CHANDLER'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'You had no relationship!!',\n","  'speaker': 'MONICA',\n","  'context': ['I just saw Danny getting on the subway with a girl and he had his arm around her.',\n","   \"Oh, honey, I'm sorry.\",\n","   'Well, you should be, this is all your fault! You meddled in our relationship!!'],\n","  'context_speakers': ['RACHEL', 'MONICA', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Well ah, I've been doing a lot of thinking, and I look at my life\",\n","  'speaker': 'PERSON',\n","  'context': ['Ooh, nice.',\n","   'So ah, there was this thing I wanted to talk to you about.',\n","   'Oh, right! I completely forgot about that.'],\n","  'context_speakers': ['MONICA', 'PERSON', 'MONICA'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"Uh, I'm gonna kill myself!\",\n","  'speaker': 'PERSON',\n","  'context': [\"I've been talking to you for five minutes now about killing myself and no one's even looked up from their desk. Hang-hang on. Hey everybody!\"],\n","  'context_speakers': ['PERSON'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': \"You and I, we're interfacing.\",\n","  'speaker': 'ROSS',\n","  'context': ['So theoretically we could download our thoughts and our memories into this computer... ...',\n","   'and live forever as a machine!',\n","   \"That's so Janine, you-you-you know what, do you know we're doing right now?\"],\n","  'context_speakers': ['ROSS', 'PERSON', 'ROSS'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'Hey.',\n","  'speaker': 'JOEY',\n","  'context': [\"Where'd you get fruit at four in the morning?\",\n","   \"Went down to the docks. Bet ya didn't know you could get it wholesale.\",\n","   \"I didn't know there were docks.\"],\n","  'context_speakers': ['ROSS', 'MONICA', 'RACHEL'],\n","  'show': 'FRIENDS',\n","  'sarcasm': False},\n"," {'utterance': 'What was I doing? My ben gazara in person?',\n","  'speaker': 'DOROTHY',\n","  'context': ['Probably just misses her mother.',\n","   'It needs to hear a feminine voice.'],\n","  'context_speakers': ['SOPHIA', 'SOPHIA'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': \"No I am upset because Joe Piscapo didn't get the lead role in rain man.\",\n","  'speaker': 'DOROTHY',\n","  'context': ['No rose.',\n","   \"Are you upset because Blanche is going to sell the house and we can't live together.\"],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': \"No, she I supset because Jimmy Swaager can't cover his motel bill\",\n","  'speaker': 'DOROTHY',\n","  'context': ['But I am still upset.',\n","   \"Are you upset because you don't know what to do about Al?\"],\n","  'context_speakers': ['ROSE', 'BLANCHE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'A hand puppet Rose.',\n","  'speaker': 'DOROTHY',\n","  'context': ['So you have replaced me.', 'With what?'],\n","  'context_speakers': ['PERSON', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'Rose, you should sit here and watch her hack herself to death',\n","  'speaker': 'DOROTHY',\n","  'context': ['ma I think you should sit down',\n","   'Dorothy, should I get Sophia a glass of water?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No Rose, that you should do.',\n","  'speaker': 'DOROTHY',\n","  'context': ['No more ice-skating',\n","   'And I am not gonna go Niagra falls in a barrel.'],\n","  'context_speakers': ['ROSE', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': \"I'm worried about whether Michael Jackson will be able to buy the remains of the elephant man.\",\n","  'speaker': 'DOROTHY',\n","  'context': ['Still worried about the money you owe the government?'],\n","  'context_speakers': ['BLANCHE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No, she is going to sit here where itÕs a 112 degrees and eat enchiladas.',\n","  'speaker': 'DOROTHY',\n","  'context': ['Its Miami in June, only cats are wearing fur!',\n","   'You going out?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No Rose, I am nervous because Sonny Bono gets elected mayor of Palm Springs, he is gonna make all the postmens wear leather bell bottoms and a fur vest.',\n","  'speaker': 'DOROTHY',\n","  'context': [\"They are ain't here yet\",\n","   'No, but they should be any minute',\n","   'You seem a little nervous.',\n","   \"Oh you're sure I am\",\n","   \"Are you nervous cause you haven't met Michael's fiancee?\"],\n","  'context_speakers': ['BLANCHE', 'DOROTHY', 'ROSE', 'DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No Rose, she is depressed because Marvel had manners on only once a week',\n","  'speaker': 'DOROTHY',\n","  'context': ['Now PERSON is depressed Rose.',\n","   'Do you think she is depressed about Rebecca'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No Blanche, with a venus fly trap',\n","  'speaker': 'DOROTHY',\n","  'context': [\"Morning everybody, Rose honey I hope you don't mind, I borrowed your golf club, I have a date to play this morning.\",\n","   'With a man?'],\n","  'context_speakers': ['DOROTHY', 'BLANCHE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No, I am depressed becaues Phillis George left the morning news.',\n","  'speaker': 'DOROTHY',\n","  'context': ['Are you depressed about Glen?'],\n","  'context_speakers': ['ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No itÕs the rose in the kitchen thing.',\n","  'speaker': 'DOROTHY',\n","  'context': ['She is miserable.',\n","   'Just to keep me upto speed Dorothy, it is this negligent arson that prospect of spending the rest of her wiening years in jail thing right?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'Oh ma no, it was a little girl selling girl scout toilets.',\n","  'speaker': 'DOROTHY',\n","  'context': ['Oh fine we will!', 'Is that Paolo?'],\n","  'context_speakers': ['DOROTHY', 'SOPHIA'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No Blanche she is upset because they keep changing the taste of coke.',\n","  'speaker': 'DOROTHY',\n","  'context': ['Honey you alright?', \"I'll be fine\", 'Is this about Arnie?'],\n","  'context_speakers': ['DOROTHY', 'ROSE', 'BLANCHE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'His cappucino maker.',\n","  'speaker': 'DOROTHY',\n","  'context': ['The only thing we have in common is under the sheets',\n","   'Whats under the sheets?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'Lords of Arabia',\n","  'speaker': 'DOROTHY',\n","  'context': ['Well ofcourse Blanche , they did it in that movie.',\n","   'What movie?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'A bucket of chicked. I hope you like it extra flat and crispy',\n","  'speaker': 'DOROTHY',\n","  'context': ['We brought dinner.', 'What you get?'],\n","  'context_speakers': ['ROSE', 'SOPHIA'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': \"The liberarche marquee at Caesar's palace\",\n","  'speaker': 'DOROTHY',\n","  'context': ['Whop, it is really coming down.', 'Whats coming down?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'The little man who lives in there, who turns the light on and off will tell you.',\n","  'speaker': 'DOROTHY',\n","  'context': ['How would I know if they fog up?'],\n","  'context_speakers': ['ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': \"It's a musical Rose\",\n","  'speaker': 'DOROTHY',\n","  'context': ['What kind of a movie is this Dorothy?'],\n","  'context_speakers': ['ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No Rose, at a pie eating contest.',\n","  'speaker': 'DOROTHY',\n","  'context': ['Oh I am so nervous. Have you ever given an eulogy?',\n","   'You mean at a funeral?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'No, its because I cannot make any sense of those commercials for obsession.',\n","  'speaker': 'DOROTHY',\n","  'context': [\"Sorry I snapped at your Rose. It's just that I am upset, I am very upset.\",\n","   'Is it because you threw Michael out?'],\n","  'context_speakers': ['DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'To feed the cat Rose.',\n","  'speaker': 'DOROTHY',\n","  'context': ['Your kidney?', 'My kidney.', 'Why would she need a kidney?'],\n","  'context_speakers': ['DOROTHY', 'BLANCHE', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'Because it wears a little sunbrella Rose.',\n","  'speaker': 'DOROTHY',\n","  'context': ['It is not a fly Rose',\n","   'Spanish fly is not a fly?',\n","   \"No, it's a beetle.\",\n","   'They call it a fly but its really a beetle?',\n","   'Yes',\n","   'How did they know its Spanish?'],\n","  'context_speakers': ['DOROTHY',\n","   'ROSE',\n","   'DOROTHY',\n","   'ROSE',\n","   'DOROTHY',\n","   'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': \"I'm sorry Sophia, but in Dorothy's defence, it sure sounds a lot like her.\",\n","  'speaker': 'ROSE',\n","  'context': [\"You what the joke going around teacher's lounge was today Ma? Me\",\n","   'I know this joke',\n","   \"Everybody heard on the radio that some anonymous Ma was complaining that her dependent daughter Dorothy doesn't have a life of her own.\",\n","   'What you think of the only grey head substitute teacher named Dorothy wasting her life away in Miami?'],\n","  'context_speakers': ['DOROTHY', 'SOPHIA', 'DOROTHY', 'SOPHIA'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'Ofcourse no Ma, I am referring to Carrey Grant, he is living in the broom closet.',\n","  'speaker': 'DOROTHY',\n","  'context': ['Supposed to study for my French final with a fourteen year old in the house, its hard enough with an eighty year old.',\n","   'Are you referring to me?'],\n","  'context_speakers': ['DOROTHY', 'SOPHIA'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'What are these? Claws?',\n","  'speaker': 'DOROTHY',\n","  'context': ['Oh Dorothy would please open this jar of nuts for me?',\n","   \"You couldn't get it open?\",\n","   \"Oh I didn't try. I don't wanna risk cracking a nail.\"],\n","  'context_speakers': ['BLANCHE', 'DOROTHY', 'BLANCHE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'In what Blanche, dog years?',\n","  'speaker': 'DOROTHY',\n","  'context': ['Was there ever any doubt?',\n","   \"Momentarily. This is strictly off the record, but Dirk's nearly five years younger that I.\"],\n","  'context_speakers': ['DOROTHY', 'BLANCHE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': \"I don't know Rose, maybe Marry Poppins has an Italian cousin.\",\n","  'speaker': 'DOROTHY',\n","  'context': [\"Isn't she happy here?\",\n","   'I think she will be happy with Gloria',\n","   'What about us? Whose gonna keep after us. Make us linguini and tell us story about Cicily.'],\n","  'context_speakers': ['ROSE', 'DOROTHY', 'ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'Show them your slides of Hawaii',\n","  'speaker': 'DOROTHY',\n","  'context': ['Wait donÕt leave me, what will I do if they come back?'],\n","  'context_speakers': ['ROSE'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': True},\n"," {'utterance': 'How about stopping by new years eve with three naked women, throwing up in the pool, passing out in the punch bowl, and relieving yourself in the linen closet?',\n","  'speaker': 'DOROTHY',\n","  'context': ['We thought you might like to join us for dinner?',\n","   'I get the feeling Dorothy that you might be a little mad at me. Why is that?'],\n","  'context_speakers': ['DOROTHY', 'SOPHIA'],\n","  'show': 'GOLDENGIRLS',\n","  'sarcasm': False},\n"," {'utterance': 'Well scott I am sure we are all really really looking forward to hear your story.',\n","  'speaker': 'MODERATOR',\n","  'context': ['I finally got to admit it to myself. I have got a sarcasm problem.'],\n","  'context_speakers': ['SCOTT'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'Sarcasm? At a sarcasmoholics meeting?',\n","  'speaker': 'MEMBER-GIRL',\n","  'context': ['Alright! Definitely!'],\n","  'context_speakers': ['SCOTT'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'No, no no, I just mediate the gorup. I just give up every Tuesday of my life to help these screwups for the good of my health',\n","  'speaker': 'MODERATOR',\n","  'context': ['Sarcasmholic yourself?'],\n","  'context_speakers': ['SCOTT'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'Noo..',\n","  'speaker': 'MODERATOR',\n","  'context': ['[nodding] - Was that sarcasm?'],\n","  'context_speakers': ['SCOTT'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'Amazing.',\n","  'speaker': 'MEMBER-GIRL',\n","  'context': ['Hello everybody! How was your week?'],\n","  'context_speakers': ['MODERATOR'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'And I am sure every person here really really wants you to join our group.',\n","  'speaker': 'MODERATOR',\n","  'context': ['Scott scott come back. You have to realize that we are a bunch of people with severe problems.',\n","   'We have been trying to help ourselves, just like you have been trying to help yourself.',\n","   \"That's why we need to help each other.\",\n","   'I mean, you seem like a decent guy.'],\n","  'context_speakers': ['MODERATOR', 'MODERATOR', 'MODERATOR', 'MODERATOR'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'Nooo!',\n","  'speaker': 'MODERATOR',\n","  'context': ['It was quite hard .. Was just being sarcastic!'],\n","  'context_speakers': ['SCOTT'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'You got off to a really good start with the group.',\n","  'speaker': 'MEMBER-BOY',\n","  'context': ['It was being sarcastic, alright I see I see',\n","   \"I didn't realise the sarcasm was going to be so full on. I ll just stand back and start again\",\n","   \"Oh, we wouldn't bother.\"],\n","  'context_speakers': ['SCOTT', 'SCOTT', 'MEMBER-BOY'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'Really?',\n","  'speaker': 'MEMBER-GIRL',\n","  'context': [\"He doesn't need to be sarcastic. I mean, that was sarcasm wasn't it?\"],\n","  'context_speakers': ['SCOTT'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True},\n"," {'utterance': 'Noooo. Really?',\n","  'speaker': 'MEMBER-BOY',\n","  'context': ['I am having a bit of trouble telling when people are sarcastic.'],\n","  'context_speakers': ['SCOTT'],\n","  'show': 'SARCASMOHOLICS',\n","  'sarcasm': True}]"]},"metadata":{},"execution_count":117}]},{"cell_type":"code","source":["train_text_input_ids, train_text_attention_mask = prepare_dataset(train_text_data)\n"],"metadata":{"id":"XQbxyXiAxNJY","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":74,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":118,"outputs":[]},{"cell_type":"code","source":["valid_text_input_ids, valid_text_attention_mask = prepare_dataset(valid_text_data)\n","test_text_input_ids, test_text_attention_mask = prepare_dataset(test_text_data)\n"],"metadata":{"id":"3qj1tJruxRC2","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":73,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":119,"outputs":[]},{"cell_type":"code","source":["print(\"TYPE : train_text_input_ids : \", type(train_text_input_ids))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"htG2YI5uxTDD","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":73,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"fec4a355-471d-4400-d767-0d515faed4b5"},"execution_count":120,"outputs":[{"output_type":"stream","name":"stdout","text":["TYPE : train_text_input_ids :  <class 'torch.Tensor'>\n"]}]},{"cell_type":"code","source":["print(tokenizer.all_special_ids)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"5eEAIwUqxWHt","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":25,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"c1198e57-3f66-4852-e8a8-15ad7dd672c7"},"execution_count":121,"outputs":[{"output_type":"stream","name":"stdout","text":["[0, 2, 3, 1, 50264, 50265, 50266]\n"]}]},{"cell_type":"code","source":["tokenizer.all_special_tokens\n"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"wg52MwbQxUXe","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":24,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"a3b8cf1c-2734-43dc-ccc9-8887fc871696"},"execution_count":122,"outputs":[{"output_type":"execute_result","data":{"text/plain":["['<s>', '</s>', '<unk>', '<pad>', '<mask>', '[CONTEXT]', '[UTTERANCE]']"]},"metadata":{},"execution_count":122}]},{"cell_type":"code","execution_count":123,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":23,"status":"ok","timestamp":1684515006955,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"oGASW03jLFtd","outputId":"df91eb14-d41d-4efd-8872-5d1185e4fba3"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["['<s>', '</s>', '<unk>', '<pad>', '<mask>', '[CONTEXT]', '[UTTERANCE]']"]},"metadata":{},"execution_count":123}],"source":["tokenizer.all_special_tokens\n"]},{"cell_type":"code","execution_count":124,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":22,"status":"ok","timestamp":1684515006955,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"JVqVQYAULBg5","outputId":"2791d4d5-e3e2-466c-f351-dd36d9ed25d6"},"outputs":[{"output_type":"stream","name":"stdout","text":["[0, 2, 3, 1, 50264, 50265, 50266]\n"]}],"source":["print(tokenizer.all_special_ids)"]},{"cell_type":"code","source":["test_label_data[0]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"278vnwZ9xZKZ","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":21,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"b797bae5-9fc8-4e8c-fffc-90b6837754f0"},"execution_count":125,"outputs":[{"output_type":"execute_result","data":{"text/plain":["['Others', 'True', -1.0, 9.0]"]},"metadata":{},"execution_count":125}]},{"cell_type":"code","source":["valid_label_data[0]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Hj1h0L_9xalM","executionInfo":{"status":"ok","timestamp":1684515006955,"user_tz":-330,"elapsed":20,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"77b9b9c6-ea1a-4bb8-f69d-eac565fe67d3"},"execution_count":126,"outputs":[{"output_type":"execute_result","data":{"text/plain":["['Others', 'True', -1.0, 6.0]"]},"metadata":{},"execution_count":126}]},{"cell_type":"code","source":["train_label_data[0]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"SylEIfjGxb8L","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":20,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"74c49873-cc56-4038-e533-ac0371ade816"},"execution_count":127,"outputs":[{"output_type":"execute_result","data":{"text/plain":["['Others', 'True', -1.0, 9.0]"]},"metadata":{},"execution_count":127}]},{"cell_type":"code","source":["label_id_dict = {'Others': 0,\n"," 'Agreement': 1,\n"," 'Statement Opinion': 2,\n"," 'Answer': 3,\n"," 'Disagreement': 4,\n"," 'Question': 5,\n"," 'Acknowledge': 6,\n"," 'Backchannel': 7,\n"," 'Statement Non Opinion': 8,\n"," 'Command': 9,\n"," 'Apology': 10,\n"," 'Greeting': 11}"],"metadata":{"id":"zWzsTR3Rxfvr","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":19,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":128,"outputs":[]},{"cell_type":"code","source":["train_intent_label = []\n","valid_intent_label = []\n","test_intent_label = []\n","\n","for j in train_label_data:\n","  # print(j)\n","  temp_intent1 = label_id_dict[j[0]]\n","  train_intent_label.append(temp_intent1)\n","\n","for k in valid_label_data:\n","  temp_intent2 = label_id_dict[k[0]]\n","  valid_intent_label.append(temp_intent2)\n","\n","for m in test_label_data:\n","  temp_intent3 = label_id_dict[m[0]]\n","  test_intent_label.append(temp_intent3)  "],"metadata":{"id":"-06QRmXqxmpW","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":18,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":129,"outputs":[]},{"cell_type":"code","source":["train_intent_label = torch.tensor(train_intent_label)\n","valid_intent_label = torch.tensor(valid_intent_label)\n","test_intent_label = torch.tensor(test_intent_label)\n","\n","print(train_intent_label.shape)\n","print(valid_intent_label.shape)\n","print(test_intent_label.shape)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Q-zvMv7oxo14","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":18,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"e27a329b-f062-4c01-b7ad-22080dcaa78f"},"execution_count":130,"outputs":[{"output_type":"stream","name":"stdout","text":["torch.Size([540])\n","torch.Size([75])\n","torch.Size([75])\n"]}]},{"cell_type":"code","source":[],"metadata":{"id":"93Cezkczy9sO","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":17,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":130,"outputs":[]},{"cell_type":"code","source":["label_id_dict_sar = {'False' : 0,\n","                 'True' : 1}"],"metadata":{"id":"Vn5FXRedyL7S","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":17,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":131,"outputs":[]},{"cell_type":"code","source":["train_sarcasm_label = []\n","valid_sarcasm_label = []\n","test_sarcasm_label = []\n","\n","for j in train_label_data:\n","  # print()\n","  temp_sarcasm1 = label_id_dict_sar[j[1]]\n","  train_sarcasm_label.append(temp_sarcasm1)\n","\n","for k in valid_label_data:\n","  temp_sarcasm2 = label_id_dict_sar[k[1]]\n","  valid_sarcasm_label.append(temp_sarcasm2)\n","\n","for u in test_label_data:\n","  temp_sarcasm3 = label_id_dict_sar[u[1]]\n","  test_sarcasm_label.append(temp_sarcasm3)"],"metadata":{"id":"JsgGJE-1yOuf","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":17,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":132,"outputs":[]},{"cell_type":"code","source":["train_sarcasm_label = torch.tensor(train_sarcasm_label)\n","valid_sarcasm_label = torch.tensor(valid_sarcasm_label)\n","test_sarcasm_label = torch.tensor(test_sarcasm_label)\n","\n","print(train_sarcasm_label.shape)\n","print(valid_sarcasm_label.shape)\n","print(test_sarcasm_label.shape)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"IkvZ7ZL8yX4q","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":17,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"b094ffde-809e-4188-f95c-b18e3b31144b"},"execution_count":133,"outputs":[{"output_type":"stream","name":"stdout","text":["torch.Size([540])\n","torch.Size([75])\n","torch.Size([75])\n"]}]},{"cell_type":"code","source":["valid_sarcasm_label"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"duc2qY1PyZsh","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":16,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"035c25a6-c1bc-48e4-cbec-5a002bf4ea1c"},"execution_count":134,"outputs":[{"output_type":"execute_result","data":{"text/plain":["tensor([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n","        0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0,\n","        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1,\n","        1, 1, 1])"]},"metadata":{},"execution_count":134}]},{"cell_type":"code","source":["test_sarcasm_label"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"AAf0FPAuybEi","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":15,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"c7a5be6e-8eb0-4dba-97c6-25e3fd0d2910"},"execution_count":135,"outputs":[{"output_type":"execute_result","data":{"text/plain":["tensor([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n","        0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0,\n","        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1,\n","        1, 1, 1])"]},"metadata":{},"execution_count":135}]},{"cell_type":"code","source":["valid_label_data[:5]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"aOVG4XR8ydEJ","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":14,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"ede6b267-d113-4b8d-b3e0-920c2a6a2ac2"},"execution_count":136,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[['Others', 'True', -1.0, 6.0],\n"," ['Others', 'True', -1.0, 4.0],\n"," ['Others', 'True', -1.0, 1.0],\n"," ['Others', 'True', -1.0, 1.0],\n"," ['Others', 'True', -1.0, 3.0]]"]},"metadata":{},"execution_count":136}]},{"cell_type":"code","source":["test_label_data[5:10]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"0-O4Ah08yjji","executionInfo":{"status":"ok","timestamp":1684515006956,"user_tz":-330,"elapsed":13,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"dc6d1ea8-3240-4161-c9d1-c33b1ac17e68"},"execution_count":137,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[['Agreement', 'True', -1.0, 4.0],\n"," ['Agreement', 'True', 1.0, 7.0],\n"," ['Answer', 'True', 1.0, 7.0],\n"," ['Answer', 'True', -1.0, 4.0],\n"," ['Answer', 'True', -1.0, 4.0]]"]},"metadata":{},"execution_count":137}]},{"cell_type":"code","source":["test_intent_label[5:10]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"TOBIELb3zER3","executionInfo":{"status":"ok","timestamp":1684515006957,"user_tz":-330,"elapsed":13,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}},"outputId":"63b5894e-3d45-412e-db7e-29f6a73f8db7"},"execution_count":138,"outputs":[{"output_type":"execute_result","data":{"text/plain":["tensor([1, 1, 3, 3, 3])"]},"metadata":{},"execution_count":138}]},{"cell_type":"code","execution_count":139,"metadata":{"id":"qrligkveIb4f","executionInfo":{"status":"ok","timestamp":1684515006957,"user_tz":-330,"elapsed":12,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class MultimodalSarcasmDataset(Dataset):\n","    # def __init__(self, utterance_input_ids, utterance_attention_mask, context_input_ids, context_attention_mask, acoustic_data, visual_data, labels):\n","    def __init__(self, utterance_input_ids, utterance_attention_mask, acoustic_data, visual_data, sarcasm_labels, intent_labels):\n","\n","        self.utterance_input_ids = utterance_input_ids\n","        self.utterance_attention_mask = utterance_attention_mask\n","        # self.context_input_ids = context_input_ids\n","        # self.context_attention_mask = context_attention_mask\n","        # self.context_attention_mask\n","        self.acoustic_data = acoustic_data\n","        self.visual_data = visual_data\n","        self.sarcasm_labels = sarcasm_labels\n","        self.intent_labels = intent_labels\n","\n","    def __len__(self):\n","        return len(self.utterance_input_ids)\n","\n","    def __getitem__(self, idx):\n","        # return self.utterance_input_ids[idx], self.utterance_attention_mask[idx], self.context_input_ids[idx], self.context_attention_mask[idx], self.acoustic_data[idx], self.visual_data[idx], self.labels[idx]\n","        return self.utterance_input_ids[idx], self.utterance_attention_mask[idx],  self.acoustic_data[idx], self.visual_data[idx], self.sarcasm_labels[idx], self.intent_labels[idx]\n","\n"]},{"cell_type":"code","execution_count":140,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":12,"status":"ok","timestamp":1684515006957,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"5j_r6ZqtIeD9","outputId":"a126dd5e-6d1f-4f98-f70d-0561a8ab5b89"},"outputs":[{"output_type":"stream","name":"stdout","text":["<torch.utils.data.dataloader.DataLoader object at 0x7f3b5c4aead0>\n"]}],"source":["\n","train_loader = DataLoader(MultimodalSarcasmDataset(train_text_input_ids, train_text_attention_mask,  train_audio_data, train_video_data, train_sarcasm_label, train_intent_label), batch_size=32, shuffle = True)\n","valid_loader = DataLoader(MultimodalSarcasmDataset(valid_text_input_ids, valid_text_attention_mask,    valid_audio_data, valid_video_data, valid_sarcasm_label, valid_intent_label), batch_size = 32, shuffle = True)\n","test_loader = DataLoader(MultimodalSarcasmDataset(test_text_input_ids, test_text_attention_mask,   test_audio_data, test_video_data, test_sarcasm_label, test_intent_label), batch_size=32, shuffle = False)\n","\n","\n","print(test_loader)\n","\n","# print(train_loader)"]},{"cell_type":"code","execution_count":141,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":11,"status":"ok","timestamp":1684515006957,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"XJYdQp5vwRim","outputId":"dedf40dc-3f9d-4d81-b247-ee262339ee88"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["540"]},"metadata":{},"execution_count":141}],"source":["len(train_loader.dataset)"]},{"cell_type":"code","execution_count":142,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":10,"status":"ok","timestamp":1684515006957,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"ZYX5DVe_wYtG","outputId":"0f147e6a-ca9a-4004-8394-bea9df9302d6"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":142}],"source":["len(valid_loader.dataset)"]},{"cell_type":"code","execution_count":143,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":9,"status":"ok","timestamp":1684515006957,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"NPTKQs5uwdsY","outputId":"72bb3e17-2104-4e9f-d7d3-69ef379b817d"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["75"]},"metadata":{},"execution_count":143}],"source":["len(test_loader.dataset)"]},{"cell_type":"code","execution_count":144,"metadata":{"id":"44TjmFmSIh9D","executionInfo":{"status":"ok","timestamp":1684515006957,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# optimizer_sarcasm_only = torch.optim.Adam(model_sarcasm_only.parameters(), lr = LEARNING_RATE)\n","# optimizer_intent_only = torch.optim.Adam(model_intent_only.parameters(), lr = LEARNING_RATE)\n","# optimizer_sarcasm_shared = torch.optim.Adam(model_intent_shared.parameters(), lr = LEARNING_RATE)\n","optimizer_intent_shared = torch.optim.Adam(model_intent_shared.parameters(), lr = LEARNING_RATE)\n","criterion = torch.nn.CrossEntropyLoss()"]},{"cell_type":"code","execution_count":145,"metadata":{"id":"x52-Vzs5p70b","executionInfo":{"status":"ok","timestamp":1684515006957,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# class label_smoothing_loss(torch.nn.Module):\n","#   def __init__(self, num_classes, smoothing = 0.1):\n","#     super(label_smoothing_loss, self).__init__()\n","#     eps = smoothing/num_classes\n","#     self.negative = eps\n","#     self.positive = (1-smoothing) + eps\n","\n","#   def forward(self, pred, target):\n","#     pred = pred.log_softmax(dim = -1)\n","#     true_dist = torch.zeros_like(pred)\n","#     true_dist.fill_(self.negative)\n","#     target_data = target.unsqueeze(dim = -1)\n","#     true_dist.scatter_(dim = -1, index  = target_data, src = torch.tensor(self.positive).expand(target_data.shape).cuda())\n","#     return torch.sum(-true_dist * pred, dim = -1).mean()   "]},{"cell_type":"code","execution_count":146,"metadata":{"id":"M5nIxIPLq76Q","executionInfo":{"status":"ok","timestamp":1684515006957,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# criterion = label_smoothing_loss(num_classes = 2, smoothing = 0.1)"]},{"cell_type":"code","execution_count":147,"metadata":{"id":"O1EC_JPPUpq_","executionInfo":{"status":"ok","timestamp":1684515006957,"user_tz":-330,"elapsed":8,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# modelsarcas/.config"]},{"cell_type":"code","execution_count":148,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":7,"status":"ok","timestamp":1684515006957,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"AbrSjHApAMiQ","outputId":"0e58bec4-e142-4f83-dd6a-a9ec9314967b"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["device(type='cuda')"]},"metadata":{},"execution_count":148}],"source":["DEVICE"]},{"cell_type":"code","execution_count":149,"metadata":{"id":"Q-tW3QhTPRtG","executionInfo":{"status":"ok","timestamp":1684515006957,"user_tz":-330,"elapsed":7,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["path = '/content/drive/MyDrive/Colab Notebooks/35/saved_model/balanced_intent/try_version/best_model_epoch_19_best_f1_36.pt'"]},{"cell_type":"code","execution_count":150,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":6161,"status":"ok","timestamp":1684515013112,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"Guhu1uW4PEkg","outputId":"bd72c759-d3ea-4730-ba29-aa40ee9a881d"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["<All keys matched successfully>"]},"metadata":{},"execution_count":150}],"source":["model_intent_only.load_state_dict(torch.load(path))"]},{"cell_type":"code","execution_count":151,"metadata":{"id":"5orEe7-pb-q-","executionInfo":{"status":"ok","timestamp":1684515013113,"user_tz":-330,"elapsed":5,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["path_sarcasm = '/content/drive/MyDrive/Colab Notebooks/35/saved_model/balanced_sarcasm/try_version/best_model_epoch_7_best_f1_72.pt'"]},{"cell_type":"code","execution_count":152,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":7098,"status":"ok","timestamp":1684515020207,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"GFhe5bUz4_ac","outputId":"86d12718-4bd5-44df-8169-52e1c94c8ff4"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["<All keys matched successfully>"]},"metadata":{},"execution_count":152}],"source":["model_sarcasm_only.load_state_dict(torch.load(path_sarcasm))"]},{"cell_type":"code","execution_count":153,"metadata":{"id":"TF3U8SkdZJ9P","executionInfo":{"status":"ok","timestamp":1684515020208,"user_tz":-330,"elapsed":15,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["model_sarcasm_only = model_sarcasm_only.to(DEVICE)\n","model_intent_only = model_intent_only.to(DEVICE)\n","\n","model_intent_shared = model_intent_shared.to(DEVICE)\n"]},{"cell_type":"code","execution_count":154,"metadata":{"id":"t98lCYDAqUlq","executionInfo":{"status":"ok","timestamp":1684515020208,"user_tz":-330,"elapsed":13,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# model_sarcasm_shared = model_sarcasm_shared.to(DEVICE)"]},{"cell_type":"code","execution_count":155,"metadata":{"id":"R8o0ielX_9n5","executionInfo":{"status":"ok","timestamp":1684515020208,"user_tz":-330,"elapsed":12,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# def train_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, model_intent_shared, data_loader):\n","def train_epoch(model_sarcasm_only, model_intent_only, model_intent_shared, data_loader):  \n","      model_sarcasm_only.eval()\n","      model_intent_only.eval()\n","      # model_sarcasm_shared.train()\n","      model_intent_shared.train()\n","\n","      epoch_train_loss = 0.0\n","    \n","      \n","      for step, batch in enumerate(tqdm(data_loader, desc = 'Training Iteration')):\n","        # for i, t in enumerate(batch):\n","        #     print(\"Inside hello\")\n","        #     print(i, \" : \", type(t))\n","        batch = tuple(t.to(DEVICE) for t in batch)\n","        # input_ids, attention_mask, context_input_ids, context_attention_mask, acoustic_input, visual_input, labels = batch\n","        input_ids, attention_mask,  acoustic_input, visual_input, sarcasm_labels, intent_labels = batch\n","        \n","        # optimizer_sarcasm_only.zero_grad()\n","        # optimizer_intent_only.zero_grad()\n","        # optimizer_sarcasm_shared.zero_grad()\n","        optimizer_intent_shared.zero_grad()\n","\n","        # print(\"Input ids shape : \", input_ids.shape)\n","        # print(\"Input ids shape : \", input_ids.shape)\n","        output_sarcasm_only, sarcasm_only_hidden_representation, sarcasm_hidden_states, sarcasm_intermediate_hidden = model_sarcasm_only(input_ids = input_ids,\n","                        attention_mask = attention_mask,\n","                        # context_input_ids = context_input_ids,\n","                        # context_attention_mask = context_attention_mask,\n","                        acoustic_input = acoustic_input,\n","                        visual_input = visual_input,\n","                        labels = sarcasm_labels)\n","        \n","        # print(output_sarcasm_only.shape)\n","        # print(sarcasm_only_hidden_representation.shape)\n","        # print(sarcasm_hidden_states.shape)\n","        \n","        output_intent_only, intent_only_hidden_representation, intent_hidden_states, intent_intermediate_hidden = model_intent_only(input_ids = input_ids,\n","                        attention_mask = attention_mask,\n","                        # context_input_ids = context_input_ids,\n","                        # context_attention_mask = context_attention_mask,\n","                        acoustic_input = acoustic_input,\n","                        visual_input = visual_input,\n","                        labels = intent_labels)\n","        \n","        # print(intent_only_hidden_representation.shape)\n","        # print(intent_hidden_states.shape)\n","        \n","        sarcasm_only_copy1 = sarcasm_only_hidden_representation.detach().clone()\n","        intent_only_copy1 = intent_only_hidden_representation.detach().clone()\n","\n","        sarcasm_only_copy2 = sarcasm_intermediate_hidden.detach().clone()\n","        intent_only_copy2 = intent_intermediate_hidden.detach().clone()\n","\n","            # output_sarcasm_shared = model_sarcasm_shared(sarcasm_only_copy, intent_only_copy)\n","            # output_sarcasm_shared = model_sarcasm_shared(sarcasm_hidden_states, intent_hidden_states, attention_mask)\n","        output_intent_shared = model_intent_shared(sarcasm_only_copy1, intent_only_copy1, sarcasm_only_copy2, intent_only_copy2)\n","\n","        # sarcasm_only_loss = output_sarcasm_only['loss']\n","        # intent_only_loss = output_intent_only['loss']\n","        # sarcasm_shared_loss = criterion(output_sarcasm_shared, sarcasm_labels)\n","        intent_shared_loss = criterion(output_intent_shared, intent_labels)\n","        \n","\n","        # loss = outputs['loss']\n","        # logits = outputs['logits']\n","        # loss = criterion(logits, labels)\n","        # epoch_train_loss += sarcasm_only_loss.item() + intent_only_loss.item() + sarcasm_shared_loss.item() + intent_shared_loss.item()\n","        # epoch_train_loss += sarcasm_only_loss.item() + intent_only_loss.item() + sarcasm_shared_loss.item() + intent_shared_loss.item()\n","        # epoch_train_loss += sarcasm_only_loss.item() +  sarcasm_shared_loss.item() \n","        # epoch_train_loss += sarcasm_shared_loss.item() \n","        epoch_train_loss +=  intent_shared_loss.item() \n","\n","        # print(\"Batch wise loss : \", epoch_train_loss)\n","\n","        # sarcasm_only_loss.backward()\n","        # intent_only_loss.backward()\n","        # sarcasm_shared_loss.backward()\n","        intent_shared_loss.backward()\n","\n","        # optimizer_sarcasm_only.step()\n","        # optimizer_intent_only.step()\n","        # optimizer_sarcasm_shared.step()\n","        optimizer_intent_shared.step()\n","\n","    \n","\n","      print(\"Epoch train loss : \", epoch_train_loss)"]},{"cell_type":"code","execution_count":156,"metadata":{"id":"RuZoFCx1ABhH","executionInfo":{"status":"ok","timestamp":1684515020209,"user_tz":-330,"elapsed":13,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# def valid_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, model_intent_shared, data_loader):\n","def valid_epoch(model_sarcasm_only, model_intent_only, model_intent_shared, data_loader):  \n","  model_sarcasm_only.eval()\n","  model_intent_only.eval()\n","  # model_sarcasm_shared.eval()\n","  model_intent_shared.eval()\n","\n","  sarcasm_only_predictions = []\n","  sarcasm_gold = []\n","  intent_only_predictions = []\n","  intent_gold = []\n","  sarcasm_shared_predictions = []\n","  intent_shared_predictions = []\n","\n","  valid_loss = 0.0\n","  with torch.no_grad():\n","    for step, batch in enumerate(tqdm(data_loader)):\n","      batch = tuple(t.to(DEVICE) for t in batch)\n","      # input_ids, attention_mask, context_input_ids, context_attention_mask, acoustic_input, visual_input, labels = batch\n","      input_ids, attention_mask,  acoustic_input, visual_input, sarcasm_labels, intent_labels = batch\n","\n","      output_sarcasm_only, sarcasm_only_hidden_representation, sarcasm_hidden_states, sarcasm_intermediate_hidden = model_sarcasm_only(input_ids = input_ids,\n","                        attention_mask = attention_mask,\n","                        # context_input_ids = context_input_ids,\n","                        # context_attention_mask = context_attention_mask,\n","                        acoustic_input = acoustic_input,\n","                        visual_input = visual_input,\n","                        labels = sarcasm_labels)\n","        \n","      output_intent_only, intent_only_hidden_representation, intent_hidden_states, intent_intermediate_hidden = model_intent_only(input_ids = input_ids,\n","                        attention_mask = attention_mask,\n","                        # context_input_ids = context_input_ids,\n","                        # context_attention_mask = context_attention_mask,\n","                        acoustic_input = acoustic_input,\n","                        visual_input = visual_input,\n","                        labels = intent_labels)\n","        \n","      sarcasm_only_copy1 = sarcasm_only_hidden_representation.detach().clone()\n","      intent_only_copy1 = intent_only_hidden_representation.detach().clone()\n","\n","      sarcasm_only_copy2 = sarcasm_intermediate_hidden.detach().clone()\n","      intent_only_copy2 = intent_intermediate_hidden.detach().clone()\n","\n","            # output_sarcasm_shared = model_sarcasm_shared(sarcasm_only_copy, intent_only_copy)\n","            # output_sarcasm_shared = model_sarcasm_shared(sarcasm_hidden_states, intent_hidden_states, attention_mask)\n","      output_intent_shared = model_intent_shared(sarcasm_only_copy1, intent_only_copy1, sarcasm_only_copy2, intent_only_copy2)\n","\n","      # sarcasm_only_loss = output_sarcasm_only['loss']\n","      # intent_only_loss = output_intent_only['loss']\n","      # sarcasm_shared_loss = criterion(output_sarcasm_shared, sarcasm_labels)\n","      intent_shared_loss = criterion(output_intent_shared, intent_labels)\n","\n","      # valid_loss += sarcasm_only_loss.item() + intent_only_loss.item() + sarcasm_shared_loss.item() + intent_shared_loss.item()\n","      # valid_loss += sarcasm_only_loss.item() + intent_only_loss.item() + sarcasm_shared_loss.item() \n","      # valid_loss += sarcasm_only_loss.item() +  sarcasm_shared_loss.item() \n","      # valid_loss += sarcasm_shared_loss.item()\n","      valid_loss += intent_shared_loss.item()\n","\n","\n","\n","      # sarcasm_only_pred = output_sarcasm_only['logits'].argmax(dim = -1)\n","      intent_only_pred = output_intent_only['logits'].argmax(dim = -1)\n","      # sarcasm_shared_pred = output_sarcasm_shared.argmax(dim = -1)\n","      intent_shared_pred = output_intent_shared.argmax(dim = -1)\n","\n","      \n","\n","      # sarcasm_only_predictions.extend(sarcasm_only_pred.tolist())\n","      intent_only_predictions.extend(intent_only_pred.tolist())\n","      # sarcasm_shared_predictions.extend(sarcasm_shared_pred.tolist())\n","      intent_shared_predictions.extend(intent_shared_pred.tolist())\n","\n","\n","      sarcasm_gold.extend(sarcasm_labels.tolist())\n","      intent_gold.extend(intent_labels.tolist())\n","\n","  return valid_loss, sarcasm_only_predictions, intent_only_predictions, sarcasm_shared_predictions, intent_shared_predictions, sarcasm_gold, intent_gold\n","\n"]},{"cell_type":"code","execution_count":157,"metadata":{"id":"FziCY7xDAEFL","executionInfo":{"status":"ok","timestamp":1684515020209,"user_tz":-330,"elapsed":12,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["\n","\n","# def test_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, model_intent_shared, data_loader):\n","def test_epoch(model_sarcasm_only, model_intent_only, model_intent_shared, data_loader):  \n","    model_sarcasm_only.eval()\n","    model_intent_only.eval()\n","    # model_sarcasm_shared.eval()\n","    model_intent_shared.eval()\n","\n","    sarcasm_only_predictions = []\n","    sarcasm_gold = []\n","    intent_only_predictions = []\n","    intent_gold = []\n","    sarcasm_shared_predictions = []\n","    intent_shared_predictions = []\n","    sarcasm_mixed_list = []\n","    intent_mixed_list = []\n","\n","    correct = 0\n","    with torch.no_grad():\n","        for step, batch in enumerate(tqdm(data_loader)):\n","            batch = tuple(t.to(DEVICE) for t in batch)\n","            # input_ids, attention_mask, context_input_ids, context_attention_mask, acoustic_input, visual_input, labels = batch\n","        \n","            input_ids, attention_mask,  acoustic_input, visual_input, sarcasm_labels, intent_labels = batch\n","\n","            output_sarcasm_only, sarcasm_only_hidden_representation, sarcasm_hidden_states, sarcasm_intermediate_hidden = model_sarcasm_only(input_ids = input_ids,\n","                              attention_mask = attention_mask,\n","                              # context_input_ids = context_input_ids,\n","                              # context_attention_mask = context_attention_mask,\n","                              acoustic_input = acoustic_input,\n","                              visual_input = visual_input,\n","                              labels = sarcasm_labels)\n","              \n","            output_intent_only, intent_only_hidden_representation, intent_hidden_states, intent_intermediate_hidden = model_intent_only(input_ids = input_ids,\n","                              attention_mask = attention_mask,\n","                              # context_input_ids = context_input_ids,\n","                              # context_attention_mask = context_attention_mask,\n","                              acoustic_input = acoustic_input,\n","                              visual_input = visual_input,\n","                              labels = intent_labels)\n","              \n","            sarcasm_only_copy1 = sarcasm_only_hidden_representation.detach().clone()\n","            intent_only_copy1 = intent_only_hidden_representation.detach().clone()\n","\n","            sarcasm_only_copy2 = sarcasm_intermediate_hidden.detach().clone()\n","            intent_only_copy2 = intent_intermediate_hidden.detach().clone()\n","\n","            # output_sarcasm_shared = model_sarcasm_shared(sarcasm_only_copy, intent_only_copy)\n","            # output_sarcasm_shared = model_sarcasm_shared(sarcasm_hidden_states, intent_hidden_states, attention_mask)\n","            output_intent_shared = model_intent_shared(sarcasm_only_copy1, intent_only_copy1, sarcasm_only_copy2, intent_only_copy2)\n","\n","\n","\n","            \n","\n","\n","            # sarcasm_only_pred = output_sarcasm_only['logits'].argmax(dim = -1)\n","            intent_only_pred = output_intent_only['logits'].argmax(dim = -1)\n","            # sarcasm_shared_pred = output_sarcasm_shared.argmax(dim = -1)\n","            intent_shared_pred = output_intent_shared.argmax(dim = -1)\n","\n","            # sarcasm_mixed = (output_sarcasm_only['logits'] + output_sarcasm_shared)/2\n","            # # print(\"sarcasm mixed shape : \", sarcasm_mixed.shape)\n","            # sarcasm_mixed_pred = sarcasm_mixed.argmax(dim = -1)\n","\n","            intent_mixed = (output_intent_only['logits'] + output_intent_shared)/2\n","            # print(\"sarcasm mixed shape : \", sarcasm_mixed.shape)\n","            intent_mixed_pred = intent_mixed.argmax(dim = -1)\n","\n","\n","            \n","\n","            # sarcasm_only_predictions.extend(sarcasm_only_pred.tolist())\n","            intent_only_predictions.extend(intent_only_pred.tolist())\n","            # sarcasm_shared_predictions.extend(sarcasm_shared_pred.tolist())\n","            intent_shared_predictions.extend(intent_shared_pred.tolist())\n","            # sarcasm_mixed_list.extend(sarcasm_mixed_pred.tolist())\n","            intent_mixed_list.extend(intent_mixed_pred.tolist())\n","\n","\n","            sarcasm_gold.extend(sarcasm_labels.tolist())\n","            intent_gold.extend(intent_labels.tolist())\n","\n","            # correct += int((pred == labels).sum())\n","\n","    return sarcasm_only_predictions, intent_only_predictions, sarcasm_shared_predictions, intent_shared_predictions, intent_mixed_list, sarcasm_gold, intent_gold\n"]},{"cell_type":"code","execution_count":158,"metadata":{"id":"s5QauBgKBtYS","executionInfo":{"status":"ok","timestamp":1684515020209,"user_tz":-330,"elapsed":11,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["class EarlyStopping:\n","  def __init__(self, patience, min_delta):\n","    self.patience = patience\n","    self.min_delta = min_delta\n","    self.counter = 0\n","    self.min_validation = np.inf\n","\n","  def early_stop(self, valid_loss):\n","    if valid_loss < self.min_validation:\n","      self.min_validation = valid_loss\n","      self.counter = 0\n","    elif valid_loss > (self.min_validation + self.min_delta):\n","      self.counter += 1\n","      if self.counter >= self.patience:\n","        return True\n","    return False          "]},{"cell_type":"code","execution_count":159,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":12,"status":"ok","timestamp":1684515020210,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"8dHmyfM2Cc9b","outputId":"c634c96f-7d18-409a-ed31-14039ad68c8e"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["<__main__.EarlyStopping at 0x7f3b5c4b7b80>"]},"metadata":{},"execution_count":159}],"source":["early_stopper = EarlyStopping(patience = 8, min_delta = 0.1)\n","early_stopper"]},{"cell_type":"code","execution_count":160,"metadata":{"id":"B0HB4PIc6ixU","executionInfo":{"status":"ok","timestamp":1684515020210,"user_tz":-330,"elapsed":10,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["\n","# def train_and_validation(model_sarcasm_only, model_intent_only, model_sarcasm_shared, model_intent_shared, train_loader, valid_loader):\n","# def train_and_validation(model_sarcasm_only, model_intent_only, model_sarcasm_shared, train_loader, valid_loader):\n","def train_and_validation(model_sarcasm_only, model_intent_only, model_intent_shared, train_loader, valid_loader):  \n","  # lowest_loss = 1e6\n","  best_f1 = 0.0\n","  # min_loss = 1e6\n","  for epoch in range(20):\n","    print(\"\\n=============Epoch : \", epoch)\n","    # train_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, model_intent_shared, train_loader)\n","    # train_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, train_loader)\n","    train_epoch(model_sarcasm_only, model_intent_only, model_intent_shared, train_loader)\n","    # valid_loss, valid_sarcasm_only_predictions, valid_intent_only_predictions, valid_sarcasm_shared_predictions, valid_intent_shared_predictions, valid_sarcasm_gold, valid_intent_gold = valid_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, model_intent_shared, valid_loader)\n","    # valid_loss, valid_sarcasm_only_predictions, valid_intent_only_predictions, valid_sarcasm_shared_predictions, valid_intent_shared_predictions, valid_sarcasm_gold, valid_intent_gold = valid_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, valid_loader)\n","\n","    valid_loss, valid_sarcasm_only_predictions, valid_intent_only_predictions, valid_sarcasm_shared_predictions, valid_intent_shared_predictions, valid_sarcasm_gold, valid_intent_gold = valid_epoch(model_sarcasm_only, model_intent_only, model_intent_shared, valid_loader)\n","\n","\n","    # if early_stopper.early_stop(valid_loss):\n","    #   break\n","\n","    # print(\"Length of predictions : \", len(valid_sarcasm_only_predictions))\n","    # print(\"Length of gold : \", len(valid_sarcasm_gold))\n","    # print(\"Valid loss : \", valid_loss)\n","    # print(\"\\n Valid Accuracy : \", accuracy_score(valid_sarcasm_gold, valid_sarcasm_only_predictions))\n","    # print(\"\\n Valid Precision : \", precision_score(valid_sarcasm_gold, valid_sarcasm_only_predictions, average = 'weighted'))\n","    # print(\"\\n Valid Recall : \", recall_score(valid_sarcasm_gold, valid_sarcasm_only_predictions, average = 'weighted'))\n","    # print(\"\\nValid F1 score : \", f1_score(valid_sarcasm_gold, valid_sarcasm_only_predictions, average = 'weighted')) \n","\n","    # print('\\n----Sarcasm shared-----\\n')\n","    # print(\"\\n Valid Accuracy : \", accuracy_score(valid_sarcasm_gold, valid_sarcasm_shared_predictions))\n","    # print(\"\\n Valid Precision : \", precision_score(valid_sarcasm_gold, valid_sarcasm_shared_predictions, average = 'weighted'))\n","    # print(\"\\n Valid Recall : \", recall_score(valid_sarcasm_gold, valid_sarcasm_shared_predictions, average = 'weighted'))\n","    # print(\"\\nValid F1 score : \", f1_score(valid_sarcasm_gold, valid_sarcasm_shared_predictions, average = 'weighted')) \n","\n","    print(\"Length of predictions : \", len(valid_intent_only_predictions))\n","    print(\"Length of gold : \", len(valid_intent_gold))\n","    print(\"Valid loss : \", valid_loss)\n","    print(\"\\n Valid Accuracy : \", accuracy_score(valid_intent_gold, valid_intent_only_predictions))\n","    print(\"\\n Valid Precision : \", precision_score(valid_intent_gold, valid_intent_only_predictions, average = 'weighted'))\n","    print(\"\\n Valid Recall : \", recall_score(valid_intent_gold, valid_intent_only_predictions, average = 'weighted'))\n","    print(\"\\nValid F1 score : \", f1_score(valid_intent_gold, valid_intent_only_predictions, average = 'weighted')) \n","\n","    print('\\n----Intent shared-----\\n')\n","    print(\"\\n Valid Accuracy : \", accuracy_score(valid_intent_gold, valid_intent_shared_predictions))\n","    print(\"\\n Valid Precision : \", precision_score(valid_intent_gold, valid_intent_shared_predictions, average = 'weighted'))\n","    print(\"\\n Valid Recall : \", recall_score(valid_intent_gold, valid_intent_shared_predictions, average = 'weighted'))\n","    print(\"\\nValid F1 score : \", f1_score(valid_intent_gold, valid_intent_shared_predictions, average = 'weighted')) \n","\n","    \n","    # curr_f1 = f1_score(valid_sarcasm_gold, valid_sarcasm_only_predictions, average = 'weighted')\n","    curr_f1 = f1_score(valid_intent_gold, valid_intent_shared_predictions, average = 'weighted')\n","\n","    curr_loss = valid_loss\n","    # if((curr_f1 > best_f1) and (epoch>=4)):\n","    if(curr_f1 > best_f1):\n","    # if(curr_loss < min_loss):    \n","    # if(curr_loss < lowest_loss):    \n","      best_f1 = curr_f1\n","      # min_loss = curr_loss\n","      # print(\"Valid pred : \", valid_pred)\n","      # print('valid_gold : ', valid_gold)\n","      torch.save(model_intent_shared.state_dict(), '/content/drive/MyDrive/Colab Notebooks/35/saved_model/balanced_intent_shared/try_version/best_model_epoch_'+str(epoch)+'_best_f1_'+str(int(best_f1*100))+'.pt')\n","      print(\"model saved\\n\")\n","\n","\n"]},{"cell_type":"code","execution_count":160,"metadata":{"id":"73CfE2olu5OS","executionInfo":{"status":"ok","timestamp":1684515020210,"user_tz":-330,"elapsed":10,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":[]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":125267,"status":"ok","timestamp":1683978181254,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"RMJL4HbkEBh2","outputId":"f7aee501-0430-4bb2-b7f6-482e798f28d3"},"outputs":[{"output_type":"stream","name":"stdout","text":["\n","=============Epoch :  0\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:08<00:00,  2.08it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  30.82624053955078\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  3.97it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  5.837627172470093\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.4\n","\n"," Valid Precision :  0.29733333333333334\n","\n"," Valid Recall :  0.4\n","\n","Valid F1 score :  0.2691216931216931\n","model saved\n","\n","\n","=============Epoch :  1\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.21it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  18.785366356372833\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  5.947383642196655\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.3466666666666667\n","\n"," Valid Precision :  0.2643270684137557\n","\n"," Valid Recall :  0.3466666666666667\n","\n","Valid F1 score :  0.2989766950551264\n","model saved\n","\n","\n","=============Epoch :  2\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  14.312902688980103\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.00it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  5.933236002922058\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.4\n","\n"," Valid Precision :  0.2893977455716586\n","\n"," Valid Recall :  0.4\n","\n","Valid F1 score :  0.3309424594873511\n","model saved\n","\n","\n","=============Epoch :  3\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  11.743258446455002\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  3.97it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  6.742872953414917\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.36\n","\n"," Valid Precision :  0.2729001080220593\n","\n"," Valid Recall :  0.36\n","\n","Valid F1 score :  0.30834138486312396\n","\n","=============Epoch :  4\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.19it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  10.083407998085022\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  3.99it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  6.473266005516052\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.37333333333333335\n","\n"," Valid Precision :  0.28044444444444444\n","\n"," Valid Recall :  0.37333333333333335\n","\n","Valid F1 score :  0.31788948306595366\n","\n","=============Epoch :  5\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  8.8618845641613\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  6.532541394233704\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.37333333333333335\n","\n"," Valid Precision :  0.29236804748999873\n","\n"," Valid Recall :  0.37333333333333335\n","\n","Valid F1 score :  0.3237826086956522\n","\n","=============Epoch :  6\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  7.788820058107376\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.00it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  7.035254001617432\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.36\n","\n"," Valid Precision :  0.2912006512006512\n","\n"," Valid Recall :  0.36\n","\n","Valid F1 score :  0.318452736318408\n","\n","=============Epoch :  7\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.19it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  6.893090099096298\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  7.791236162185669\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.37333333333333335\n","\n"," Valid Precision :  0.3177001945422998\n","\n"," Valid Recall :  0.37333333333333335\n","\n","Valid F1 score :  0.3388095447225882\n","model saved\n","\n","\n","=============Epoch :  8\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.21it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  6.1350478529930115\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  7.6236937046051025\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.4\n","\n"," Valid Precision :  0.34530771372876634\n","\n"," Valid Recall :  0.4\n","\n","Valid F1 score :  0.36463636363636365\n","model saved\n","\n","\n","=============Epoch :  9\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.23it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  5.489215984940529\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.253284215927124\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.36323483923483924\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.3754576787236634\n","model saved\n","\n","\n","=============Epoch :  10\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.22it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  4.9008506536483765\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.749183177947998\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.36\n","\n"," Valid Precision :  0.35351611351611356\n","\n"," Valid Recall :  0.36\n","\n","Valid F1 score :  0.3487470725995316\n","\n","=============Epoch :  11\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  4.400344684720039\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.02it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  7.15148651599884\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.38666666666666666\n","\n"," Valid Precision :  0.3671896511896512\n","\n"," Valid Recall :  0.38666666666666666\n","\n","Valid F1 score :  0.3656164807059948\n","\n","=============Epoch :  12\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.21it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  3.963938996195793\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.071569442749023\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.38666666666666666\n","\n"," Valid Precision :  0.3686210826210827\n","\n"," Valid Recall :  0.38666666666666666\n","\n","Valid F1 score :  0.3667630008525149\n","\n","=============Epoch :  13\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  3.616406589746475\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  3.99it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.735815286636353\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.37333333333333335\n","\n"," Valid Precision :  0.3548381574852163\n","\n"," Valid Recall :  0.37333333333333335\n","\n","Valid F1 score :  0.35489058253574385\n","\n","=============Epoch :  14\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  3.201063856482506\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.161534070968628\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.4\n","\n"," Valid Precision :  0.3707663707663708\n","\n"," Valid Recall :  0.4\n","\n","Valid F1 score :  0.37544561235865587\n","\n","=============Epoch :  15\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  2.880574844777584\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  3.99it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.067513465881348\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.36\n","\n"," Valid Precision :  0.3403679653679653\n","\n"," Valid Recall :  0.36\n","\n","Valid F1 score :  0.3452391360915951\n","\n","=============Epoch :  16\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.22it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  2.7377541810274124\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  3.99it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.806574821472168\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.38666666666666666\n","\n"," Valid Precision :  0.35695622895622897\n","\n"," Valid Recall :  0.38666666666666666\n","\n","Valid F1 score :  0.35864705882352943\n","\n","=============Epoch :  17\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  2.418903484940529\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.00it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.681144118309021\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.37333333333333335\n","\n"," Valid Precision :  0.3656228956228956\n","\n"," Valid Recall :  0.37333333333333335\n","\n","Valid F1 score :  0.3606065573770492\n","\n","=============Epoch :  18\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.20it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  2.1630382239818573\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  8.451487302780151\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.36\n","\n"," Valid Precision :  0.3454949494949495\n","\n"," Valid Recall :  0.36\n","\n","Valid F1 score :  0.34720365523258484\n","\n","=============Epoch :  19\n"]},{"output_type":"stream","name":"stderr","text":["Training Iteration: 100%|██████████| 17/17 [00:05<00:00,  3.19it/s]\n"]},{"output_type":"stream","name":"stdout","text":["Epoch train loss :  1.983190167695284\n"]},{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:00<00:00,  4.01it/s]"]},{"output_type":"stream","name":"stdout","text":["Length of predictions :  75\n","Length of gold :  75\n","Valid loss :  10.127203702926636\n","\n"," Valid Accuracy :  0.41333333333333333\n","\n"," Valid Precision :  0.35821082621082617\n","\n"," Valid Recall :  0.41333333333333333\n","\n","Valid F1 score :  0.36263026375210944\n","\n","----Intent shared-----\n","\n","\n"," Valid Accuracy :  0.38666666666666666\n","\n"," Valid Precision :  0.36822668822668825\n","\n"," Valid Recall :  0.38666666666666666\n","\n","Valid F1 score :  0.36630275521579875\n"]},{"output_type":"stream","name":"stderr","text":["\n"]}],"source":["# train_and_validation(model, train_loader, test_loader)\n","# train_and_validation(model_sarcasm_only, model_intent_only, model_sarcasm_shared, model_intent_shared, train_loader, valid_loader)\n","# train_and_validation(model_sarcasm_only, model_intent_only, model_sarcasm_shared, train_loader, valid_loader)\n","# train_and_validation(model_sarcasm_only, model_intent_only, model_sarcasm_shared, train_loader, test_loader)\n","# train_and_validation(model_sarcasm_only, train_loader, valid_loader)\n","train_and_validation(model_sarcasm_only, model_intent_only, model_intent_shared, train_loader, valid_loader)"]},{"cell_type":"code","source":[],"metadata":{"id":"RrLH9aihhKQ4"},"execution_count":null,"outputs":[]},{"cell_type":"code","execution_count":161,"metadata":{"id":"DO0Qsq9ufvhi","executionInfo":{"status":"ok","timestamp":1684515036503,"user_tz":-330,"elapsed":3,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"outputs":[],"source":["# path = '/content/drive/MyDrive/Colab Notebooks/32/saved_model_f1/stratify/best_model_epoch_11_best_f1_78_foldNum_0.pt'\n","\n","path = '/content/drive/MyDrive/Colab Notebooks/35/saved_model/balanced_intent_shared/try_version/best_model_epoch_9_best_f1_37.pt'"]},{"cell_type":"code","execution_count":162,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":1109,"status":"ok","timestamp":1684515037610,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"_J_qDFj5_ie3","outputId":"1f9e3490-222c-464f-caf7-dfc36b80342c"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["<All keys matched successfully>"]},"metadata":{},"execution_count":162}],"source":["# model_sarcasm_shared.load_state_dict(torch.load(path))\n","model_intent_shared.load_state_dict(torch.load(path))"]},{"cell_type":"code","execution_count":163,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":4340,"status":"ok","timestamp":1684515042526,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"},"user_tz":-330},"id":"UgEpzAt-_mSQ","outputId":"16df21e1-6095-4763-ad4c-047b43d83bd5"},"outputs":[{"output_type":"stream","name":"stderr","text":["100%|██████████| 3/3 [00:04<00:00,  1.40s/it]"]},{"output_type":"stream","name":"stdout","text":["\n","Accuracy :  0.4266666666666667\n","\n","Precision :  0.3213037037037037\n","\n","Recall :  0.4266666666666667\n","\n","F1 score :  0.3536935286935287\n","\n","==========\n","\n","\n","Accuracy :  0.48\n","\n","Precision :  0.38684702093397744\n","\n","Recall :  0.48\n","\n","F1 score :  0.4120473414591062\n","\n","Confusion Matrix : \n"," [[22  0  1  4  0  0  0  1  0  0  0  0]\n"," [ 4  1  1  0  0  0  0  0  0  0  0  0]\n"," [ 4  0  1  1  0  1  0  0  0  0  0  0]\n"," [ 3  0  0  5  0  1  0  0  1  0  0  0]\n"," [ 4  1  0  0  0  0  0  0  0  0  0  0]\n"," [ 1  0  0  0  0  7  0  0  0  0  0  0]\n"," [ 0  0  1  0  0  0  0  0  0  0  0  0]\n"," [ 3  0  0  0  0  0  0  0  0  0  0  0]\n"," [ 3  0  0  0  0  0  0  0  0  0  0  0]\n"," [ 1  0  1  0  0  0  0  0  0  0  0  0]\n"," [ 0  0  0  0  0  0  0  0  1  0  0  0]\n"," [ 1  0  0  0  0  0  0  0  0  0  0  0]]\n"]},{"output_type":"stream","name":"stderr","text":["\n"]}],"source":["\n","\n","# test_sarcasm_only_predictions, test_intent_only_predictions, test_sarcasm_shared_predictions, test_intent_shared_predictions, test_sarcasm_mixed, test_sarcasm_gold, intent_gold = test_epoch(model_sarcasm_only, model_intent_only, model_sarcasm_shared, test_loader)\n","\n","test_sarcasm_only_predictions, test_intent_only_predictions, test_sarcasm_shared_predictions, test_intent_shared_predictions, test_intent_mixed, test_sarcasm_gold, test_intent_gold = test_epoch(model_sarcasm_only, model_intent_only, model_intent_shared, test_loader)\n","\n","\n","# print(acc)\n","\n","print(\"\\nAccuracy : \", accuracy_score(test_intent_gold, test_intent_only_predictions))\n","print(\"\\nPrecision : \", precision_score(test_intent_gold, test_intent_only_predictions, average = 'weighted'))\n","print(\"\\nRecall : \", recall_score(test_intent_gold, test_intent_only_predictions, average = 'weighted'))\n","print(\"\\nF1 score : \", f1_score(test_intent_gold, test_intent_only_predictions, average = 'weighted'))\n","\n","print('\\n==========\\n')\n","\n","print(\"\\nAccuracy : \", accuracy_score(test_intent_gold, test_intent_shared_predictions))\n","print(\"\\nPrecision : \", precision_score(test_intent_gold, test_intent_shared_predictions, average = 'weighted'))\n","print(\"\\nRecall : \", recall_score(test_intent_gold, test_intent_shared_predictions, average = 'weighted'))\n","print(\"\\nF1 score : \", f1_score(test_intent_gold, test_intent_shared_predictions, average = 'weighted'))\n","\n","# print('\\n==========\\n')\n","\n","\n","# print(\"\\nAccuracy : \", accuracy_score(test_intent_gold, test_intent_mixed))\n","# print(\"\\nPrecision : \", precision_score(test_intent_gold, test_intent_mixed, average = 'weighted'))\n","# print(\"\\nRecall : \", recall_score(test_intent_gold, test_intent_mixed, average = 'weighted'))\n","# print(\"\\nF1 score : \", f1_score(test_intent_gold, test_intent_mixed, average = 'weighted'))\n","        \n","\n","\n","\n","\n","print('\\nConfusion Matrix : \\n', confusion_matrix(test_intent_gold, test_intent_shared_predictions))\n","\n","# sar = 0\n","\n","# for j in range(len(test_pred)):\n","#   if((test_pred[j]==1) and (test_gold[j] == 1)):\n","#     sar+=1\n","\n","\n","# print(\"Sarcastic count : \", sar)\n"]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/saved_model/output_file/test_intent_multitask_gold.p', 'wb') as f:\n","  pickle.dump(test_intent_gold, f)"],"metadata":{"id":"BG7Yhipg6wlz","executionInfo":{"status":"ok","timestamp":1684515051078,"user_tz":-330,"elapsed":5,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":164,"outputs":[]},{"cell_type":"code","source":["with open('/content/drive/MyDrive/Colab Notebooks/35/saved_model/output_file/test_intent_multitask_pred.p', 'wb') as f:\n","  pickle.dump(test_intent_shared_predictions, f)"],"metadata":{"id":"fgHvp-MvlLQL","executionInfo":{"status":"ok","timestamp":1684515051079,"user_tz":-330,"elapsed":4,"user":{"displayName":"Mohit Singh Tomar","userId":"08465857202059996456"}}},"execution_count":165,"outputs":[]},{"cell_type":"code","source":[],"metadata":{"id":"pibn27K5lLj-"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":[],"metadata":{"id":"InTBlsGDlL5Y"},"execution_count":null,"outputs":[]}],"metadata":{"accelerator":"GPU","colab":{"machine_shape":"hm","provenance":[]},"gpuClass":"premium","kernelspec":{"display_name":"Python 3","name":"python3"},"language_info":{"name":"python"},"widgets":{"application/vnd.jupyter.widget-state+json":{"8ebacbe3dc344455befd54fc5665fd1b":{"model_module":"@jupyter-widgets/controls","model_name":"HBoxModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_3133a56799144a108de38093072999cd","IPY_MODEL_a7e0f241ef744f4bae89846070aac64b","IPY_MODEL_0ae280d726d444bbae665c7fb4994a29"],"layout":"IPY_MODEL_a928128d4e364b6dbc40e45acdc5fdd5"}},"3133a56799144a108de38093072999cd":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_2004beaac327433a88c0cc186b6da8be","placeholder":"​","style":"IPY_MODEL_00d79bb4d88e45ce8a2361ca3eda44bb","value":"Downloading (…)lve/main/config.json: 100%"}},"a7e0f241ef744f4bae89846070aac64b":{"model_module":"@jupyter-widgets/controls","model_name":"FloatProgressModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_58085460ac1e479581252e38ff3b5340","max":1716,"min":0,"orientation":"horizontal","style":"IPY_MODEL_93db01d807dd4250b3c14d3746ce14a2","value":1716}},"0ae280d726d444bbae665c7fb4994a29":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_3b3ddba2572c42b5ab5e6da9e19e9570","placeholder":"​","style":"IPY_MODEL_89b3c334d2cf4b89b23932456c012a54","value":" 1.72k/1.72k [00:00&lt;00:00, 136kB/s]"}},"a928128d4e364b6dbc40e45acdc5fdd5":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"2004beaac327433a88c0cc186b6da8be":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"00d79bb4d88e45ce8a2361ca3eda44bb":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"58085460ac1e479581252e38ff3b5340":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"93db01d807dd4250b3c14d3746ce14a2":{"model_module":"@jupyter-widgets/controls","model_name":"ProgressStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"3b3ddba2572c42b5ab5e6da9e19e9570":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"89b3c334d2cf4b89b23932456c012a54":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"dfdc980705b746898f6db4a14086b5df":{"model_module":"@jupyter-widgets/controls","model_name":"HBoxModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_7e4680003db447f6972e46f8e9bf3ce4","IPY_MODEL_bb530047087042c79a8af0b97c6c19ed","IPY_MODEL_012dacacfdcf4ec1b8b13106d4e65bae"],"layout":"IPY_MODEL_17066910932d4524a4b8fbb25e0a54e2"}},"7e4680003db447f6972e46f8e9bf3ce4":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_30f1f20b18524603923227792608a355","placeholder":"​","style":"IPY_MODEL_4d24131933004629a69c0fd75bcb7ec3","value":"Downloading pytorch_model.bin: 100%"}},"bb530047087042c79a8af0b97c6c19ed":{"model_module":"@jupyter-widgets/controls","model_name":"FloatProgressModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_0e4f53830ba9473eaa7346c253d24c06","max":557771387,"min":0,"orientation":"horizontal","style":"IPY_MODEL_ab656a90064e41889c53c640acb00e4a","value":557771387}},"012dacacfdcf4ec1b8b13106d4e65bae":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_625701156ac74b7fa66a5dd857495ed0","placeholder":"​","style":"IPY_MODEL_69214f05eee74a9dacf3bf3dd42ed458","value":" 558M/558M [00:02&lt;00:00, 247MB/s]"}},"17066910932d4524a4b8fbb25e0a54e2":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"30f1f20b18524603923227792608a355":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"4d24131933004629a69c0fd75bcb7ec3":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"0e4f53830ba9473eaa7346c253d24c06":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ab656a90064e41889c53c640acb00e4a":{"model_module":"@jupyter-widgets/controls","model_name":"ProgressStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"625701156ac74b7fa66a5dd857495ed0":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"69214f05eee74a9dacf3bf3dd42ed458":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"e266f9a9eb5b481f9be0b0f407073db6":{"model_module":"@jupyter-widgets/controls","model_name":"HBoxModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_e06b31e87af449ed90dc814b2d5168a4","IPY_MODEL_c89ca5e444654d0ba11faf028a06a57f","IPY_MODEL_796e8c3bee6b4da294ca53c4253fc189"],"layout":"IPY_MODEL_1450831a48954df4a26a7d3248e2a9bd"}},"e06b31e87af449ed90dc814b2d5168a4":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_c301f9b100ad4d2f902072a7100afadb","placeholder":"​","style":"IPY_MODEL_c1ea5bbf73bf4405898bd93120e40a41","value":"Downloading (…)olve/main/vocab.json: 100%"}},"c89ca5e444654d0ba11faf028a06a57f":{"model_module":"@jupyter-widgets/controls","model_name":"FloatProgressModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_d7833fcf43b345fd8b31de3d4d4c640f","max":898823,"min":0,"orientation":"horizontal","style":"IPY_MODEL_ca554554f6594d17b4470c270f8f894a","value":898823}},"796e8c3bee6b4da294ca53c4253fc189":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_6b32ead6e79a433d9f5355efd2bf68e6","placeholder":"​","style":"IPY_MODEL_b3e3079a882f43c9adcb9c1e9b51997c","value":" 899k/899k [00:00&lt;00:00, 13.3MB/s]"}},"1450831a48954df4a26a7d3248e2a9bd":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"c301f9b100ad4d2f902072a7100afadb":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"c1ea5bbf73bf4405898bd93120e40a41":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d7833fcf43b345fd8b31de3d4d4c640f":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ca554554f6594d17b4470c270f8f894a":{"model_module":"@jupyter-widgets/controls","model_name":"ProgressStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"6b32ead6e79a433d9f5355efd2bf68e6":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b3e3079a882f43c9adcb9c1e9b51997c":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"cb2ae3714da14a86a6fac3931bc5fa8b":{"model_module":"@jupyter-widgets/controls","model_name":"HBoxModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_f20d3416e5b34a3ea1b574dcc3e242e6","IPY_MODEL_f52a8877833f4157b873d07aa36c057d","IPY_MODEL_0b24ad5aa31f4ee5ab95a8476397af1c"],"layout":"IPY_MODEL_ef45571d08a24746afaca14cd1a17a4e"}},"f20d3416e5b34a3ea1b574dcc3e242e6":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_27437b8101444ef5a355450c67e6af25","placeholder":"​","style":"IPY_MODEL_dc64c16ebc9a465b8d847a522f70cdb2","value":"Downloading (…)olve/main/merges.txt: 100%"}},"f52a8877833f4157b873d07aa36c057d":{"model_module":"@jupyter-widgets/controls","model_name":"FloatProgressModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_d6e9277a98e74805ad6cd54936e18f77","max":456318,"min":0,"orientation":"horizontal","style":"IPY_MODEL_02a9ae0e473d4ff59c897d41fa03dcda","value":456318}},"0b24ad5aa31f4ee5ab95a8476397af1c":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_083c3a66fba34396bd1a0e50e4bed84d","placeholder":"​","style":"IPY_MODEL_6eee378f1c7b45d0b9cc79b5776ddaea","value":" 456k/456k [00:00&lt;00:00, 26.7MB/s]"}},"ef45571d08a24746afaca14cd1a17a4e":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"27437b8101444ef5a355450c67e6af25":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"dc64c16ebc9a465b8d847a522f70cdb2":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d6e9277a98e74805ad6cd54936e18f77":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"02a9ae0e473d4ff59c897d41fa03dcda":{"model_module":"@jupyter-widgets/controls","model_name":"ProgressStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"083c3a66fba34396bd1a0e50e4bed84d":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"6eee378f1c7b45d0b9cc79b5776ddaea":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"391921869d734be5adc96f25de92e793":{"model_module":"@jupyter-widgets/controls","model_name":"HBoxModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_dbb378ddfc914fa98b47e4d0e3a37be6","IPY_MODEL_5fb909faa6924c0ea137eb9d5419144b","IPY_MODEL_d79c10248bb94161bdbdf15f9e76fe76"],"layout":"IPY_MODEL_3da0e0b652224cbb9414afbab0c235ba"}},"dbb378ddfc914fa98b47e4d0e3a37be6":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_0a555dcee82046898f35ce03e4fb7ab4","placeholder":"​","style":"IPY_MODEL_5164a93bf0ac411782e2ca32d4a9742a","value":"Downloading (…)/main/tokenizer.json: 100%"}},"5fb909faa6924c0ea137eb9d5419144b":{"model_module":"@jupyter-widgets/controls","model_name":"FloatProgressModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_0b7fbae40e5a4a299ab28540c3a4083f","max":1355863,"min":0,"orientation":"horizontal","style":"IPY_MODEL_a4a111be2f264f7992f28427270ed9b5","value":1355863}},"d79c10248bb94161bdbdf15f9e76fe76":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_28c2cab9b929414496329971a1e564bc","placeholder":"​","style":"IPY_MODEL_8591e8ab909641098c641ef244bbbce2","value":" 1.36M/1.36M [00:00&lt;00:00, 50.1MB/s]"}},"3da0e0b652224cbb9414afbab0c235ba":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"0a555dcee82046898f35ce03e4fb7ab4":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"5164a93bf0ac411782e2ca32d4a9742a":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"0b7fbae40e5a4a299ab28540c3a4083f":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a4a111be2f264f7992f28427270ed9b5":{"model_module":"@jupyter-widgets/controls","model_name":"ProgressStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"28c2cab9b929414496329971a1e564bc":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"8591e8ab909641098c641ef244bbbce2":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}}}}},"nbformat":4,"nbformat_minor":0}