latentspace-plot.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5213ba2b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ['CUDA_VISIBLE_DEVICES'] = '0'\n",
    "os.environ['CUDA_PATH'] = '/usr/local/cuda-10.0'\n",
    "\n",
    "import transformers\n",
    "import datasets\n",
    "\n",
    "import torch\n",
    "from pathlib import Path\n",
    "from datasets import load_dataset\n",
    "from transformers import (AutoTokenizer, AutoModelForSequenceClassification, \n",
    "                          TrainingArguments, Trainer)\n",
    "\n",
    "import numpy as np\n",
    "import re\n",
    "import pandas as pd\n",
    "from tqdm.notebook import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e7b1b620",
   "metadata": {},
   "outputs": [],
   "source": [
    "form = 'red2'\n",
    "label_len = 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "46d3c16d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using custom data configuration default-d9ecd716484d21f3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading and preparing dataset json/default to /home/14cfl/.cache/huggingface/datasets/json/default-d9ecd716484d21f3/0.0.0/ac0ca5f5289a6cf108e706efcf040422dbbfa8e658dee6a819f20d76bb84d26b...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3e46424642834568a5e398446b88ae80",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading data files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "556a0d63b00446ee8fe9345e83b3e7d3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Extracting data files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using custom data configuration default-d9ecd716484d21f3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset json downloaded and prepared to /home/14cfl/.cache/huggingface/datasets/json/default-d9ecd716484d21f3/0.0.0/ac0ca5f5289a6cf108e706efcf040422dbbfa8e658dee6a819f20d76bb84d26b. Subsequent calls will reuse this data.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reusing dataset json (/home/14cfl/.cache/huggingface/datasets/json/default-d9ecd716484d21f3/0.0.0/ac0ca5f5289a6cf108e706efcf040422dbbfa8e658dee6a819f20d76bb84d26b)\n",
      "Using custom data configuration default-f0e098c8b2d88652\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading and preparing dataset json/default to /home/14cfl/.cache/huggingface/datasets/json/default-f0e098c8b2d88652/0.0.0/ac0ca5f5289a6cf108e706efcf040422dbbfa8e658dee6a819f20d76bb84d26b...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "75c47e9c3a144ee18109c60a0565719d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading data files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d3bd68fb0abd4b77ab5114ce63c97329",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Extracting data files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset json downloaded and prepared to /home/14cfl/.cache/huggingface/datasets/json/default-f0e098c8b2d88652/0.0.0/ac0ca5f5289a6cf108e706efcf040422dbbfa8e658dee6a819f20d76bb84d26b. Subsequent calls will reuse this data.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using custom data configuration default-1a49b4eb58afb15c\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading and preparing dataset json/default to /home/14cfl/.cache/huggingface/datasets/json/default-1a49b4eb58afb15c/0.0.0/ac0ca5f5289a6cf108e706efcf040422dbbfa8e658dee6a819f20d76bb84d26b...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d8b8595c388e4a87815c9e16626c1fc7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading data files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b2a82288f81a4be092cf35f2da0d6c12",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Extracting data files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset json downloaded and prepared to /home/14cfl/.cache/huggingface/datasets/json/default-1a49b4eb58afb15c/0.0.0/ac0ca5f5289a6cf108e706efcf040422dbbfa8e658dee6a819f20d76bb84d26b. Subsequent calls will reuse this data.\n"
     ]
    }
   ],
   "source": [
    "test_enc_1 = load_dataset('json', data_files=f'data/{form}/testenc-{label_len}.json', split='train[:50%]')\n",
    "test_enc_2 = load_dataset('json', data_files=f'data/{form}/testenc-{label_len}.json', split='train[50%:]')\n",
    "val_enc = load_dataset('json', data_files=f'data/{form}/valenc-{label_len}.json', split='train')\n",
    "train_enc = load_dataset('json', data_files=f'data/{form}/train-citationcleaned-{label_len}.json', split='train')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3845847c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['files', 'labels', 'input', 'original', 'codes']\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8b63d2e92f594ec8ad8b62cf0d322f3d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/142 [00:00<?, ?ba/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "model_ckpt = 'baseline/baseline-512-20-nlpaueb/legal-bert-base-uncased/checkpoint-34000/'\n",
    "\n",
    "tokenizer = AutoTokenizer.from_pretrained(model_ckpt)\n",
    "\n",
    "def tokenize_and_encode_in(examples):\n",
    "    return tokenizer(examples['input'], max_length=512, truncation=True, padding=True)\n",
    "\n",
    "cols = train_enc.column_names\n",
    "print(cols)\n",
    "cols.remove(\"labels\")\n",
    "train_enc = train_enc.map(tokenize_and_encode_in, batched=True, remove_columns=cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "9a767a97",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "================================================================================\n",
       "Layer (type:depth-idx)                                  Param #\n",
       "================================================================================\n",
       "BertForSequenceClassification                           --\n",
       "├─BertModel: 1-1                                        --\n",
       "│    └─BertEmbeddings: 2-1                              --\n",
       "│    │    └─Embedding: 3-1                              23,440,896\n",
       "│    │    └─Embedding: 3-2                              393,216\n",
       "│    │    └─Embedding: 3-3                              1,536\n",
       "│    │    └─LayerNorm: 3-4                              1,536\n",
       "│    │    └─Dropout: 3-5                                --\n",
       "│    └─BertEncoder: 2-2                                 --\n",
       "│    │    └─ModuleList: 3-6                             85,054,464\n",
       "│    └─BertPooler: 2-3                                  --\n",
       "│    │    └─Linear: 3-7                                 590,592\n",
       "│    │    └─Tanh: 3-8                                   --\n",
       "├─Dropout: 1-2                                          --\n",
       "├─Linear: 1-3                                           15,380\n",
       "================================================================================\n",
       "Total params: 109,497,620\n",
       "Trainable params: 109,497,620\n",
       "Non-trainable params: 0\n",
       "================================================================================"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from transformers import AutoModel,AutoModelForSequenceClassification, AdamW, get_linear_schedule_with_warmup\n",
    "from torchinfo import summary\n",
    "device = 'cuda'\n",
    "model = AutoModelForSequenceClassification.from_pretrained(model_ckpt).to(device)\n",
    "\n",
    "#out = model(sample, )\n",
    "summary(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "41d947ca",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Defaulting to user installation because normal site-packages is not writeable\n",
      "Requirement already satisfied: matplotlib in /home/14cfl/.local/lib/python3.6/site-packages (3.3.4)\n",
      "Requirement already satisfied: numpy>=1.15 in /home/14cfl/.local/lib/python3.6/site-packages (from matplotlib) (1.19.5)\n",
      "Requirement already satisfied: python-dateutil>=2.1 in /home/14cfl/.local/lib/python3.6/site-packages (from matplotlib) (2.8.2)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /home/14cfl/.local/lib/python3.6/site-packages (from matplotlib) (8.4.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /home/14cfl/.local/lib/python3.6/site-packages (from matplotlib) (1.3.1)\n",
      "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.3 in /home/14cfl/.local/lib/python3.6/site-packages (from matplotlib) (3.0.9)\n",
      "Requirement already satisfied: cycler>=0.10 in /home/14cfl/.local/lib/python3.6/site-packages (from matplotlib) (0.11.0)\n",
      "Requirement already satisfied: six>=1.5 in /home/14cfl/.local/lib/python3.6/site-packages (from python-dateutil>=2.1->matplotlib) (1.16.0)\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    }
   ],
   "source": [
    "pip install matplotlib"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8f59a201",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/14cfl/.local/lib/python3.6/site-packages/numba/core/errors.py:154: UserWarning: Insufficiently recent colorama version found. Numba requires colorama >= 0.3.9\n",
      "  warnings.warn(msg)\n"
     ]
    }
   ],
   "source": [
    "import umap\n",
    "\n",
    "from sklearn.manifold import TSNE\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "dim_reduc = umap.UMAP(random_state=10, metric='euclidean')\n",
    "\n",
    "def plot_umap(normalized_df, colour, name, prototypes=None, definitions=None):\n",
    "    standard_embedding = dim_reduc.fit_transform(normalized_df)\n",
    "    plt.scatter(standard_embedding[:, 0], standard_embedding[:, 1], c=colour, cmap='YlGnBu', s=1)\n",
    "    \n",
    "    if prototypes is not None:\n",
    "        prot_embedding = dim_reduc.transform(prototypes)\n",
    "\n",
    "        plt.scatter(prot_embedding[:, 0], prot_embedding[:, 1], color='r', s=5)\n",
    "    \n",
    "    if definitions is not None:\n",
    "        def_embedding = dim_reduc.transform(definitions)\n",
    "        plt.scatter(def_embedding[:, 0], def_embedding[:, 1], color='m', s=5)\n",
    "    \n",
    "    plt.savefig(name, dpi=200, bb_inches='tight')\n",
    "    plt.show()\n",
    "\n",
    "def plot_tsne(normalized_df, colour):\n",
    "    standard_embedding = TSNE(n_components=2, perplexity=8, learning_rate='auto',init='random').fit_transform(normalized_df)\n",
    "    plt.scatter(standard_embedding[:, 0], standard_embedding[:, 1], c=colour, s=1, cmap='bwr')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7d5aa692",
   "metadata": {},
   "outputs": [],
   "source": [
    "dl = torch.utils.data.DataLoader(train_enc, batch_size=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "25df74c7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a4b8b3bcf8f84c66940cd0c4cd725207",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "obtaining hidden states:   0%|          | 0/17655 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "plotting data points\n"
     ]
    }
   ],
   "source": [
    "# print(model.num_labels, model.device)\n",
    "device = model.device\n",
    "#if state.epoch >= 6 and state.epoch % 5 == 0:\n",
    "model.eval()\n",
    "x = []\n",
    "samples = [[] for i in range(label_len)]\n",
    "# p_per_class = int(/label_len)\n",
    "\n",
    "label = []\n",
    "\n",
    "with torch.no_grad():\n",
    "    for data in tqdm(dl, desc=\"obtaining hidden states\"):\n",
    "        # encoding data to hidden state - dim should be (N, D) - N batch size, D hidden vector dim\n",
    "        inp = torch.t(torch.stack(data['input_ids']))\n",
    "        mask = torch.t(torch.stack(data['attention_mask']))\n",
    "#         print(inp[:,:512].shape, mask.shape)\n",
    "        labels = torch.t(torch.stack(data.pop(\"labels\")))\n",
    "        outputs = model(input_ids=inp.to(device), attention_mask=mask.to(device))\n",
    "#         print(dir(outputs))\n",
    "        logits = outputs.logits.cpu()\n",
    "        #print(logits.shape)\n",
    "        # reversing one-hot encoding of the labels\n",
    "        labs = (labels == 1).nonzero()\n",
    "\n",
    "        #print(labs)\n",
    "        labels = {}\n",
    "        for l in labs:\n",
    "            if l[0].item() in labels:\n",
    "                labels[l[0].item()].append(l[1].item())\n",
    "            else:\n",
    "                labels[l[0].item()] = [l[1].item()]\n",
    "        #print(labels)\n",
    "        # adding each hidden state to its respective class' list\n",
    "        for k, v in labels.items():\n",
    "            for c in v:\n",
    "                #print(c)\n",
    "                x.append(logits[k, :])\n",
    "                label.append(c)\n",
    "                samples[c].append(inp[k,:])\n",
    "    print('plotting data points')\n",
    "    #print([len(x) for x in x_per_class])\n",
    "#     # projecting prototypes for each class\n",
    "#     print('prototypes:')\n",
    "#     for i in range(model.prototype_sample.data.size(0)):\n",
    "#         for j in range(model.prototype_sample.data.size(1)):\n",
    "#             with open(f'prototypes-red2-20-defs/_{i}_{j}.txt', 'w') as f:\n",
    "#                 text = tokenizer.decode(model.prototype_sample.data[i,j,:].int())\n",
    "#                 print(text)\n",
    "#                 f.write(text)\n",
    "\n",
    "    x = np.stack(x)\n",
    "    labels = np.stack(label)\n",
    "    \n",
    "# #     prots = model.prototype_vectors.cpu()\n",
    "#     defs = model.roberta(model.definition_tokens).last_hidden_state[:,0,:].cpu()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "69af537d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/14cfl/.local/lib/python3.6/site-packages/ipykernel_launcher.py:21: MatplotlibDeprecationWarning: savefig() got unexpected keyword argument \"bb_inches\" which is no longer supported as of 3.3 and will become an error two minor releases later\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plot_umap(x, labels, 'legalbert-finetuned-20.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6c48c425",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}