2064 lines
69 KiB
Plaintext
2064 lines
69 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "82f9eacb",
|
||
"metadata": {
|
||
"origin_pos": 0
|
||
},
|
||
"source": [
|
||
"# 用于预训练词嵌入的数据集\n",
|
||
":label:`sec_word2vec_data`\n",
|
||
"\n",
|
||
"现在我们已经了解了word2vec模型的技术细节和大致的训练方法，让我们来看看它们的实现。具体地说，我们将以 :numref:`sec_word2vec`的跳元模型和 :numref:`sec_approx_train`的负采样为例。本节从用于预训练词嵌入模型的数据集开始：数据的原始格式将被转换为可以在训练期间迭代的小批量。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 1,
|
||
"id": "596ed133",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:38.933299Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:38.932361Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:41.929964Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:41.928691Z"
|
||
},
|
||
"origin_pos": 2,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"import math\n",
|
||
"import os\n",
|
||
"import random\n",
|
||
"import torch\n",
|
||
"from d2l import torch as d2l"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "8286adf0",
|
||
"metadata": {
|
||
"origin_pos": 4
|
||
},
|
||
"source": [
|
||
"## 读取数据集\n",
|
||
"\n",
|
||
"我们在这里使用的数据集是[Penn Tree Bank（PTB）](https://catalog.ldc.upenn.edu/LDC99T42)。该语料库取自“华尔街日报”的文章，分为训练集、验证集和测试集。在原始格式中，文本文件的每一行表示由空格分隔的一句话。在这里，我们将每个单词视为一个词元。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 2,
|
||
"id": "cc6c9b2e",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:41.935897Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:41.934975Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:42.345380Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:42.344041Z"
|
||
},
|
||
"origin_pos": 5,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Downloading ../data/ptb.zip from http://d2l-data.s3-accelerate.amazonaws.com/ptb.zip...\n"
|
||
]
|
||
},
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"'# sentences数: 42069'"
|
||
]
|
||
},
|
||
"execution_count": 2,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"#@save\n",
|
||
"d2l.DATA_HUB['ptb'] = (d2l.DATA_URL + 'ptb.zip',\n",
|
||
" '319d85e578af0cdc590547f26231e4e31cdf1e42')\n",
|
||
"\n",
|
||
"#@save\n",
|
||
"def read_ptb():\n",
|
||
" \"\"\"将PTB数据集加载到文本行的列表中\"\"\"\n",
|
||
" data_dir = d2l.download_extract('ptb')\n",
|
||
" # Readthetrainingset.\n",
|
||
" with open(os.path.join(data_dir, 'ptb.train.txt')) as f:\n",
|
||
" raw_text = f.read()\n",
|
||
" return [line.split() for line in raw_text.split('\\n')]\n",
|
||
"\n",
|
||
"sentences = read_ptb()\n",
|
||
"f'# sentences数: {len(sentences)}'"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "e7290de5",
|
||
"metadata": {
|
||
"origin_pos": 6
|
||
},
|
||
"source": [
|
||
"在读取训练集之后，我们为语料库构建了一个词表，其中出现次数少于10次的任何单词都将由“<unk>”词元替换。请注意，原始数据集还包含表示稀有（未知）单词的“<unk>”词元。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 3,
|
||
"id": "04285c2d",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:42.350103Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:42.349586Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:42.520737Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:42.519523Z"
|
||
},
|
||
"origin_pos": 7,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"'vocab size: 6719'"
|
||
]
|
||
},
|
||
"execution_count": 3,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"vocab = d2l.Vocab(sentences, min_freq=10)\n",
|
||
"f'vocab size: {len(vocab)}'"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "0bba2291",
|
||
"metadata": {
|
||
"origin_pos": 8
|
||
},
|
||
"source": [
|
||
"## 下采样\n",
|
||
"\n",
|
||
"文本数据通常有“the”“a”和“in”等高频词：它们在非常大的语料库中甚至可能出现数十亿次。然而，这些词经常在上下文窗口中与许多不同的词共同出现，提供的有用信息很少。例如，考虑上下文窗口中的词“chip”：直观地说，它与低频单词“intel”的共现比与高频单词“a”的共现在训练中更有用。此外，大量（高频）单词的训练速度很慢。因此，当训练词嵌入模型时，可以对高频单词进行*下采样* :cite:`Mikolov.Sutskever.Chen.ea.2013`。具体地说，数据集中的每个词$w_i$将有概率地被丢弃\n",
|
||
"\n",
|
||
"$$ P(w_i) = \\max\\left(1 - \\sqrt{\\frac{t}{f(w_i)}}, 0\\right),$$\n",
|
||
"\n",
|
||
"其中$f(w_i)$是$w_i$的词数与数据集中的总词数的比率，常量$t$是超参数（在实验中为$10^{-4}$）。我们可以看到，只有当相对比率$f(w_i) > t$时，（高频）词$w_i$才能被丢弃，且该词的相对比率越高，被丢弃的概率就越大。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 4,
|
||
"id": "88d0f9c2",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:42.524901Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:42.524245Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:44.019122Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:44.017912Z"
|
||
},
|
||
"origin_pos": 9,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"#@save\n",
|
||
"def subsample(sentences, vocab):\n",
|
||
" \"\"\"下采样高频词\"\"\"\n",
|
||
" # 排除未知词元'<unk>'\n",
|
||
" sentences = [[token for token in line if vocab[token] != vocab.unk]\n",
|
||
" for line in sentences]\n",
|
||
" counter = d2l.count_corpus(sentences)\n",
|
||
" num_tokens = sum(counter.values())\n",
|
||
"\n",
|
||
" # 如果在下采样期间保留词元，则返回True\n",
|
||
" def keep(token):\n",
|
||
" return(random.uniform(0, 1) <\n",
|
||
" math.sqrt(1e-4 / counter[token] * num_tokens))\n",
|
||
"\n",
|
||
" return ([[token for token in line if keep(token)] for line in sentences],\n",
|
||
" counter)\n",
|
||
"\n",
|
||
"subsampled, counter = subsample(sentences, vocab)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "5c892ade",
|
||
"metadata": {
|
||
"origin_pos": 10
|
||
},
|
||
"source": [
|
||
"下面的代码片段绘制了下采样前后每句话的词元数量的直方图。正如预期的那样，下采样通过删除高频词来显著缩短句子，这将使训练加速。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 5,
|
||
"id": "5dd0b4f6",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:44.024294Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:44.023765Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:44.272889Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:44.271933Z"
|
||
},
|
||
"origin_pos": 11,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"image/svg+xml": [
|
||
"<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
|
||
"<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
|
||
" \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
|
||
"<svg xmlns:xlink=\"http://www.w3.org/1999/xlink\" width=\"262.190625pt\" height=\"182.053046pt\" viewBox=\"0 0 262.190625 182.053046\" xmlns=\"http://www.w3.org/2000/svg\" version=\"1.1\">\n",
|
||
" <metadata>\n",
|
||
" <rdf:RDF xmlns:dc=\"http://purl.org/dc/elements/1.1/\" xmlns:cc=\"http://creativecommons.org/ns#\" xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\">\n",
|
||
" <cc:Work>\n",
|
||
" <dc:type rdf:resource=\"http://purl.org/dc/dcmitype/StillImage\"/>\n",
|
||
" <dc:date>2023-08-18T07:01:44.224820</dc:date>\n",
|
||
" <dc:format>image/svg+xml</dc:format>\n",
|
||
" <dc:creator>\n",
|
||
" <cc:Agent>\n",
|
||
" <dc:title>Matplotlib v3.5.1, https://matplotlib.org/</dc:title>\n",
|
||
" </cc:Agent>\n",
|
||
" </dc:creator>\n",
|
||
" </cc:Work>\n",
|
||
" </rdf:RDF>\n",
|
||
" </metadata>\n",
|
||
" <defs>\n",
|
||
" <style type=\"text/css\">*{stroke-linejoin: round; stroke-linecap: butt}</style>\n",
|
||
" </defs>\n",
|
||
" <g id=\"figure_1\">\n",
|
||
" <g id=\"patch_1\">\n",
|
||
" <path d=\"M 0 182.053046 \n",
|
||
"L 262.190625 182.053046 \n",
|
||
"L 262.190625 0 \n",
|
||
"L 0 0 \n",
|
||
"L 0 182.053046 \n",
|
||
"z\n",
|
||
"\" style=\"fill: none\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"axes_1\">\n",
|
||
" <g id=\"patch_2\">\n",
|
||
" <path d=\"M 59.690625 144.496796 \n",
|
||
"L 254.990625 144.496796 \n",
|
||
"L 254.990625 8.596796 \n",
|
||
"L 59.690625 8.596796 \n",
|
||
"z\n",
|
||
"\" style=\"fill: #ffffff\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_3\">\n",
|
||
" <path d=\"M 68.567898 144.496796 \n",
|
||
"L 75.814651 144.496796 \n",
|
||
"L 75.814651 123.324081 \n",
|
||
"L 68.567898 123.324081 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_4\">\n",
|
||
" <path d=\"M 86.684781 144.496796 \n",
|
||
"L 93.931534 144.496796 \n",
|
||
"L 93.931534 85.795242 \n",
|
||
"L 86.684781 85.795242 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_5\">\n",
|
||
" <path d=\"M 104.801664 144.496796 \n",
|
||
"L 112.048417 144.496796 \n",
|
||
"L 112.048417 76.680027 \n",
|
||
"L 104.801664 76.680027 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_6\">\n",
|
||
" <path d=\"M 122.918547 144.496796 \n",
|
||
"L 130.1653 144.496796 \n",
|
||
"L 130.1653 96.491067 \n",
|
||
"L 122.918547 96.491067 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_7\">\n",
|
||
" <path d=\"M 141.03543 144.496796 \n",
|
||
"L 148.282183 144.496796 \n",
|
||
"L 148.282183 124.48284 \n",
|
||
"L 141.03543 124.48284 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_8\">\n",
|
||
" <path d=\"M 159.152313 144.496796 \n",
|
||
"L 166.399067 144.496796 \n",
|
||
"L 166.399067 137.6991 \n",
|
||
"L 159.152313 137.6991 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_9\">\n",
|
||
" <path d=\"M 177.269196 144.496796 \n",
|
||
"L 184.51595 144.496796 \n",
|
||
"L 184.51595 142.958904 \n",
|
||
"L 177.269196 142.958904 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_10\">\n",
|
||
" <path d=\"M 195.38608 144.496796 \n",
|
||
"L 202.632833 144.496796 \n",
|
||
"L 202.632833 144.106983 \n",
|
||
"L 195.38608 144.106983 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_11\">\n",
|
||
" <path d=\"M 213.502963 144.496796 \n",
|
||
"L 220.749716 144.496796 \n",
|
||
"L 220.749716 144.363299 \n",
|
||
"L 213.502963 144.363299 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_12\">\n",
|
||
" <path d=\"M 231.619846 144.496796 \n",
|
||
"L 238.866599 144.496796 \n",
|
||
"L 238.866599 144.422038 \n",
|
||
"L 231.619846 144.422038 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_13\">\n",
|
||
" <path d=\"M 75.814651 144.496796 \n",
|
||
"L 83.061404 144.496796 \n",
|
||
"L 83.061404 15.068225 \n",
|
||
"L 75.814651 15.068225 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_14\">\n",
|
||
" <path d=\"M 93.931534 144.496796 \n",
|
||
"L 101.178287 144.496796 \n",
|
||
"L 101.178287 59.469519 \n",
|
||
"L 93.931534 59.469519 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_15\">\n",
|
||
" <path d=\"M 112.048417 144.496796 \n",
|
||
"L 119.29517 144.496796 \n",
|
||
"L 119.29517 134.91701 \n",
|
||
"L 112.048417 134.91701 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_16\">\n",
|
||
" <path d=\"M 130.1653 144.496796 \n",
|
||
"L 137.412054 144.496796 \n",
|
||
"L 137.412054 143.941446 \n",
|
||
"L 130.1653 143.941446 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_17\">\n",
|
||
" <path d=\"M 148.282183 144.496796 \n",
|
||
"L 155.528937 144.496796 \n",
|
||
"L 155.528937 144.448737 \n",
|
||
"L 148.282183 144.448737 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_18\">\n",
|
||
" <path d=\"M 166.399067 144.496796 \n",
|
||
"L 173.64582 144.496796 \n",
|
||
"L 173.64582 144.491456 \n",
|
||
"L 166.399067 144.491456 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_19\">\n",
|
||
" <path d=\"M 184.51595 144.496796 \n",
|
||
"L 191.762703 144.496796 \n",
|
||
"L 191.762703 144.496796 \n",
|
||
"L 184.51595 144.496796 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_20\">\n",
|
||
" <path d=\"M 202.632833 144.496796 \n",
|
||
"L 209.879586 144.496796 \n",
|
||
"L 209.879586 144.496796 \n",
|
||
"L 202.632833 144.496796 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_21\">\n",
|
||
" <path d=\"M 220.749716 144.496796 \n",
|
||
"L 227.996469 144.496796 \n",
|
||
"L 227.996469 144.496796 \n",
|
||
"L 220.749716 144.496796 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_22\">\n",
|
||
" <path d=\"M 238.866599 144.496796 \n",
|
||
"L 246.113352 144.496796 \n",
|
||
"L 246.113352 144.496796 \n",
|
||
"L 238.866599 144.496796 \n",
|
||
"z\n",
|
||
"\" clip-path=\"url(#p224370b0b9)\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"matplotlib.axis_1\">\n",
|
||
" <g id=\"xtick_1\">\n",
|
||
" <g id=\"line2d_1\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"mb08744926d\" d=\"M 0 0 \n",
|
||
"L 0 3.5 \n",
|
||
"\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </defs>\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#mb08744926d\" x=\"66.756209\" y=\"144.496796\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_1\">\n",
|
||
" <!-- 0 -->\n",
|
||
" <g transform=\"translate(63.574959 159.095234)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-30\" d=\"M 2034 4250 \n",
|
||
"Q 1547 4250 1301 3770 \n",
|
||
"Q 1056 3291 1056 2328 \n",
|
||
"Q 1056 1369 1301 889 \n",
|
||
"Q 1547 409 2034 409 \n",
|
||
"Q 2525 409 2770 889 \n",
|
||
"Q 3016 1369 3016 2328 \n",
|
||
"Q 3016 3291 2770 3770 \n",
|
||
"Q 2525 4250 2034 4250 \n",
|
||
"z\n",
|
||
"M 2034 4750 \n",
|
||
"Q 2819 4750 3233 4129 \n",
|
||
"Q 3647 3509 3647 2328 \n",
|
||
"Q 3647 1150 3233 529 \n",
|
||
"Q 2819 -91 2034 -91 \n",
|
||
"Q 1250 -91 836 529 \n",
|
||
"Q 422 1150 422 2328 \n",
|
||
"Q 422 3509 836 4129 \n",
|
||
"Q 1250 4750 2034 4750 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"xtick_2\">\n",
|
||
" <g id=\"line2d_2\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#mb08744926d\" x=\"110.943729\" y=\"144.496796\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_2\">\n",
|
||
" <!-- 20 -->\n",
|
||
" <g transform=\"translate(104.581229 159.095234)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-32\" d=\"M 1228 531 \n",
|
||
"L 3431 531 \n",
|
||
"L 3431 0 \n",
|
||
"L 469 0 \n",
|
||
"L 469 531 \n",
|
||
"Q 828 903 1448 1529 \n",
|
||
"Q 2069 2156 2228 2338 \n",
|
||
"Q 2531 2678 2651 2914 \n",
|
||
"Q 2772 3150 2772 3378 \n",
|
||
"Q 2772 3750 2511 3984 \n",
|
||
"Q 2250 4219 1831 4219 \n",
|
||
"Q 1534 4219 1204 4116 \n",
|
||
"Q 875 4013 500 3803 \n",
|
||
"L 500 4441 \n",
|
||
"Q 881 4594 1212 4672 \n",
|
||
"Q 1544 4750 1819 4750 \n",
|
||
"Q 2544 4750 2975 4387 \n",
|
||
"Q 3406 4025 3406 3419 \n",
|
||
"Q 3406 3131 3298 2873 \n",
|
||
"Q 3191 2616 2906 2266 \n",
|
||
"Q 2828 2175 2409 1742 \n",
|
||
"Q 1991 1309 1228 531 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-32\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"63.623047\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"xtick_3\">\n",
|
||
" <g id=\"line2d_3\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#mb08744926d\" x=\"155.131249\" y=\"144.496796\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_3\">\n",
|
||
" <!-- 40 -->\n",
|
||
" <g transform=\"translate(148.768749 159.095234)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-34\" d=\"M 2419 4116 \n",
|
||
"L 825 1625 \n",
|
||
"L 2419 1625 \n",
|
||
"L 2419 4116 \n",
|
||
"z\n",
|
||
"M 2253 4666 \n",
|
||
"L 3047 4666 \n",
|
||
"L 3047 1625 \n",
|
||
"L 3713 1625 \n",
|
||
"L 3713 1100 \n",
|
||
"L 3047 1100 \n",
|
||
"L 3047 0 \n",
|
||
"L 2419 0 \n",
|
||
"L 2419 1100 \n",
|
||
"L 313 1100 \n",
|
||
"L 313 1709 \n",
|
||
"L 2253 4666 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-34\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"63.623047\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"xtick_4\">\n",
|
||
" <g id=\"line2d_4\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#mb08744926d\" x=\"199.318769\" y=\"144.496796\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_4\">\n",
|
||
" <!-- 60 -->\n",
|
||
" <g transform=\"translate(192.956269 159.095234)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-36\" d=\"M 2113 2584 \n",
|
||
"Q 1688 2584 1439 2293 \n",
|
||
"Q 1191 2003 1191 1497 \n",
|
||
"Q 1191 994 1439 701 \n",
|
||
"Q 1688 409 2113 409 \n",
|
||
"Q 2538 409 2786 701 \n",
|
||
"Q 3034 994 3034 1497 \n",
|
||
"Q 3034 2003 2786 2293 \n",
|
||
"Q 2538 2584 2113 2584 \n",
|
||
"z\n",
|
||
"M 3366 4563 \n",
|
||
"L 3366 3988 \n",
|
||
"Q 3128 4100 2886 4159 \n",
|
||
"Q 2644 4219 2406 4219 \n",
|
||
"Q 1781 4219 1451 3797 \n",
|
||
"Q 1122 3375 1075 2522 \n",
|
||
"Q 1259 2794 1537 2939 \n",
|
||
"Q 1816 3084 2150 3084 \n",
|
||
"Q 2853 3084 3261 2657 \n",
|
||
"Q 3669 2231 3669 1497 \n",
|
||
"Q 3669 778 3244 343 \n",
|
||
"Q 2819 -91 2113 -91 \n",
|
||
"Q 1303 -91 875 529 \n",
|
||
"Q 447 1150 447 2328 \n",
|
||
"Q 447 3434 972 4092 \n",
|
||
"Q 1497 4750 2381 4750 \n",
|
||
"Q 2619 4750 2861 4703 \n",
|
||
"Q 3103 4656 3366 4563 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-36\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"63.623047\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"xtick_5\">\n",
|
||
" <g id=\"line2d_5\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#mb08744926d\" x=\"243.506289\" y=\"144.496796\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_5\">\n",
|
||
" <!-- 80 -->\n",
|
||
" <g transform=\"translate(237.143789 159.095234)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-38\" d=\"M 2034 2216 \n",
|
||
"Q 1584 2216 1326 1975 \n",
|
||
"Q 1069 1734 1069 1313 \n",
|
||
"Q 1069 891 1326 650 \n",
|
||
"Q 1584 409 2034 409 \n",
|
||
"Q 2484 409 2743 651 \n",
|
||
"Q 3003 894 3003 1313 \n",
|
||
"Q 3003 1734 2745 1975 \n",
|
||
"Q 2488 2216 2034 2216 \n",
|
||
"z\n",
|
||
"M 1403 2484 \n",
|
||
"Q 997 2584 770 2862 \n",
|
||
"Q 544 3141 544 3541 \n",
|
||
"Q 544 4100 942 4425 \n",
|
||
"Q 1341 4750 2034 4750 \n",
|
||
"Q 2731 4750 3128 4425 \n",
|
||
"Q 3525 4100 3525 3541 \n",
|
||
"Q 3525 3141 3298 2862 \n",
|
||
"Q 3072 2584 2669 2484 \n",
|
||
"Q 3125 2378 3379 2068 \n",
|
||
"Q 3634 1759 3634 1313 \n",
|
||
"Q 3634 634 3220 271 \n",
|
||
"Q 2806 -91 2034 -91 \n",
|
||
"Q 1263 -91 848 271 \n",
|
||
"Q 434 634 434 1313 \n",
|
||
"Q 434 1759 690 2068 \n",
|
||
"Q 947 2378 1403 2484 \n",
|
||
"z\n",
|
||
"M 1172 3481 \n",
|
||
"Q 1172 3119 1398 2916 \n",
|
||
"Q 1625 2713 2034 2713 \n",
|
||
"Q 2441 2713 2670 2916 \n",
|
||
"Q 2900 3119 2900 3481 \n",
|
||
"Q 2900 3844 2670 4047 \n",
|
||
"Q 2441 4250 2034 4250 \n",
|
||
"Q 1625 4250 1398 4047 \n",
|
||
"Q 1172 3844 1172 3481 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-38\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"63.623047\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_6\">\n",
|
||
" <!-- # tokens per sentence -->\n",
|
||
" <g transform=\"translate(100.6125 172.773359)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-23\" d=\"M 3272 2816 \n",
|
||
"L 2363 2816 \n",
|
||
"L 2100 1772 \n",
|
||
"L 3016 1772 \n",
|
||
"L 3272 2816 \n",
|
||
"z\n",
|
||
"M 2803 4594 \n",
|
||
"L 2478 3297 \n",
|
||
"L 3391 3297 \n",
|
||
"L 3719 4594 \n",
|
||
"L 4219 4594 \n",
|
||
"L 3897 3297 \n",
|
||
"L 4872 3297 \n",
|
||
"L 4872 2816 \n",
|
||
"L 3775 2816 \n",
|
||
"L 3519 1772 \n",
|
||
"L 4513 1772 \n",
|
||
"L 4513 1294 \n",
|
||
"L 3397 1294 \n",
|
||
"L 3072 0 \n",
|
||
"L 2572 0 \n",
|
||
"L 2894 1294 \n",
|
||
"L 1978 1294 \n",
|
||
"L 1656 0 \n",
|
||
"L 1153 0 \n",
|
||
"L 1478 1294 \n",
|
||
"L 494 1294 \n",
|
||
"L 494 1772 \n",
|
||
"L 1594 1772 \n",
|
||
"L 1856 2816 \n",
|
||
"L 850 2816 \n",
|
||
"L 850 3297 \n",
|
||
"L 1978 3297 \n",
|
||
"L 2297 4594 \n",
|
||
"L 2803 4594 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-20\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-74\" d=\"M 1172 4494 \n",
|
||
"L 1172 3500 \n",
|
||
"L 2356 3500 \n",
|
||
"L 2356 3053 \n",
|
||
"L 1172 3053 \n",
|
||
"L 1172 1153 \n",
|
||
"Q 1172 725 1289 603 \n",
|
||
"Q 1406 481 1766 481 \n",
|
||
"L 2356 481 \n",
|
||
"L 2356 0 \n",
|
||
"L 1766 0 \n",
|
||
"Q 1100 0 847 248 \n",
|
||
"Q 594 497 594 1153 \n",
|
||
"L 594 3053 \n",
|
||
"L 172 3053 \n",
|
||
"L 172 3500 \n",
|
||
"L 594 3500 \n",
|
||
"L 594 4494 \n",
|
||
"L 1172 4494 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-6f\" d=\"M 1959 3097 \n",
|
||
"Q 1497 3097 1228 2736 \n",
|
||
"Q 959 2375 959 1747 \n",
|
||
"Q 959 1119 1226 758 \n",
|
||
"Q 1494 397 1959 397 \n",
|
||
"Q 2419 397 2687 759 \n",
|
||
"Q 2956 1122 2956 1747 \n",
|
||
"Q 2956 2369 2687 2733 \n",
|
||
"Q 2419 3097 1959 3097 \n",
|
||
"z\n",
|
||
"M 1959 3584 \n",
|
||
"Q 2709 3584 3137 3096 \n",
|
||
"Q 3566 2609 3566 1747 \n",
|
||
"Q 3566 888 3137 398 \n",
|
||
"Q 2709 -91 1959 -91 \n",
|
||
"Q 1206 -91 779 398 \n",
|
||
"Q 353 888 353 1747 \n",
|
||
"Q 353 2609 779 3096 \n",
|
||
"Q 1206 3584 1959 3584 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-6b\" d=\"M 581 4863 \n",
|
||
"L 1159 4863 \n",
|
||
"L 1159 1991 \n",
|
||
"L 2875 3500 \n",
|
||
"L 3609 3500 \n",
|
||
"L 1753 1863 \n",
|
||
"L 3688 0 \n",
|
||
"L 2938 0 \n",
|
||
"L 1159 1709 \n",
|
||
"L 1159 0 \n",
|
||
"L 581 0 \n",
|
||
"L 581 4863 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-65\" d=\"M 3597 1894 \n",
|
||
"L 3597 1613 \n",
|
||
"L 953 1613 \n",
|
||
"Q 991 1019 1311 708 \n",
|
||
"Q 1631 397 2203 397 \n",
|
||
"Q 2534 397 2845 478 \n",
|
||
"Q 3156 559 3463 722 \n",
|
||
"L 3463 178 \n",
|
||
"Q 3153 47 2828 -22 \n",
|
||
"Q 2503 -91 2169 -91 \n",
|
||
"Q 1331 -91 842 396 \n",
|
||
"Q 353 884 353 1716 \n",
|
||
"Q 353 2575 817 3079 \n",
|
||
"Q 1281 3584 2069 3584 \n",
|
||
"Q 2775 3584 3186 3129 \n",
|
||
"Q 3597 2675 3597 1894 \n",
|
||
"z\n",
|
||
"M 3022 2063 \n",
|
||
"Q 3016 2534 2758 2815 \n",
|
||
"Q 2500 3097 2075 3097 \n",
|
||
"Q 1594 3097 1305 2825 \n",
|
||
"Q 1016 2553 972 2059 \n",
|
||
"L 3022 2063 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-6e\" d=\"M 3513 2113 \n",
|
||
"L 3513 0 \n",
|
||
"L 2938 0 \n",
|
||
"L 2938 2094 \n",
|
||
"Q 2938 2591 2744 2837 \n",
|
||
"Q 2550 3084 2163 3084 \n",
|
||
"Q 1697 3084 1428 2787 \n",
|
||
"Q 1159 2491 1159 1978 \n",
|
||
"L 1159 0 \n",
|
||
"L 581 0 \n",
|
||
"L 581 3500 \n",
|
||
"L 1159 3500 \n",
|
||
"L 1159 2956 \n",
|
||
"Q 1366 3272 1645 3428 \n",
|
||
"Q 1925 3584 2291 3584 \n",
|
||
"Q 2894 3584 3203 3211 \n",
|
||
"Q 3513 2838 3513 2113 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-73\" d=\"M 2834 3397 \n",
|
||
"L 2834 2853 \n",
|
||
"Q 2591 2978 2328 3040 \n",
|
||
"Q 2066 3103 1784 3103 \n",
|
||
"Q 1356 3103 1142 2972 \n",
|
||
"Q 928 2841 928 2578 \n",
|
||
"Q 928 2378 1081 2264 \n",
|
||
"Q 1234 2150 1697 2047 \n",
|
||
"L 1894 2003 \n",
|
||
"Q 2506 1872 2764 1633 \n",
|
||
"Q 3022 1394 3022 966 \n",
|
||
"Q 3022 478 2636 193 \n",
|
||
"Q 2250 -91 1575 -91 \n",
|
||
"Q 1294 -91 989 -36 \n",
|
||
"Q 684 19 347 128 \n",
|
||
"L 347 722 \n",
|
||
"Q 666 556 975 473 \n",
|
||
"Q 1284 391 1588 391 \n",
|
||
"Q 1994 391 2212 530 \n",
|
||
"Q 2431 669 2431 922 \n",
|
||
"Q 2431 1156 2273 1281 \n",
|
||
"Q 2116 1406 1581 1522 \n",
|
||
"L 1381 1569 \n",
|
||
"Q 847 1681 609 1914 \n",
|
||
"Q 372 2147 372 2553 \n",
|
||
"Q 372 3047 722 3315 \n",
|
||
"Q 1072 3584 1716 3584 \n",
|
||
"Q 2034 3584 2315 3537 \n",
|
||
"Q 2597 3491 2834 3397 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-70\" d=\"M 1159 525 \n",
|
||
"L 1159 -1331 \n",
|
||
"L 581 -1331 \n",
|
||
"L 581 3500 \n",
|
||
"L 1159 3500 \n",
|
||
"L 1159 2969 \n",
|
||
"Q 1341 3281 1617 3432 \n",
|
||
"Q 1894 3584 2278 3584 \n",
|
||
"Q 2916 3584 3314 3078 \n",
|
||
"Q 3713 2572 3713 1747 \n",
|
||
"Q 3713 922 3314 415 \n",
|
||
"Q 2916 -91 2278 -91 \n",
|
||
"Q 1894 -91 1617 61 \n",
|
||
"Q 1341 213 1159 525 \n",
|
||
"z\n",
|
||
"M 3116 1747 \n",
|
||
"Q 3116 2381 2855 2742 \n",
|
||
"Q 2594 3103 2138 3103 \n",
|
||
"Q 1681 3103 1420 2742 \n",
|
||
"Q 1159 2381 1159 1747 \n",
|
||
"Q 1159 1113 1420 752 \n",
|
||
"Q 1681 391 2138 391 \n",
|
||
"Q 2594 391 2855 752 \n",
|
||
"Q 3116 1113 3116 1747 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-72\" d=\"M 2631 2963 \n",
|
||
"Q 2534 3019 2420 3045 \n",
|
||
"Q 2306 3072 2169 3072 \n",
|
||
"Q 1681 3072 1420 2755 \n",
|
||
"Q 1159 2438 1159 1844 \n",
|
||
"L 1159 0 \n",
|
||
"L 581 0 \n",
|
||
"L 581 3500 \n",
|
||
"L 1159 3500 \n",
|
||
"L 1159 2956 \n",
|
||
"Q 1341 3275 1631 3429 \n",
|
||
"Q 1922 3584 2338 3584 \n",
|
||
"Q 2397 3584 2469 3576 \n",
|
||
"Q 2541 3569 2628 3553 \n",
|
||
"L 2631 2963 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-63\" d=\"M 3122 3366 \n",
|
||
"L 3122 2828 \n",
|
||
"Q 2878 2963 2633 3030 \n",
|
||
"Q 2388 3097 2138 3097 \n",
|
||
"Q 1578 3097 1268 2742 \n",
|
||
"Q 959 2388 959 1747 \n",
|
||
"Q 959 1106 1268 751 \n",
|
||
"Q 1578 397 2138 397 \n",
|
||
"Q 2388 397 2633 464 \n",
|
||
"Q 2878 531 3122 666 \n",
|
||
"L 3122 134 \n",
|
||
"Q 2881 22 2623 -34 \n",
|
||
"Q 2366 -91 2075 -91 \n",
|
||
"Q 1284 -91 818 406 \n",
|
||
"Q 353 903 353 1747 \n",
|
||
"Q 353 2603 823 3093 \n",
|
||
"Q 1294 3584 2113 3584 \n",
|
||
"Q 2378 3584 2631 3529 \n",
|
||
"Q 2884 3475 3122 3366 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-23\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-20\" x=\"83.789062\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-74\" x=\"115.576172\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6f\" x=\"154.785156\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6b\" x=\"215.966797\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-65\" x=\"270.251953\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6e\" x=\"331.775391\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-73\" x=\"395.154297\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-20\" x=\"447.253906\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-70\" x=\"479.041016\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-65\" x=\"542.517578\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-72\" x=\"604.041016\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-20\" x=\"645.154297\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-73\" x=\"676.941406\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-65\" x=\"729.041016\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6e\" x=\"790.564453\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-74\" x=\"853.943359\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-65\" x=\"893.152344\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6e\" x=\"954.675781\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-63\" x=\"1018.054688\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-65\" x=\"1073.035156\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"matplotlib.axis_2\">\n",
|
||
" <g id=\"ytick_1\">\n",
|
||
" <g id=\"line2d_6\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"m7f47d3da76\" d=\"M 0 0 \n",
|
||
"L -3.5 0 \n",
|
||
"\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </defs>\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#m7f47d3da76\" x=\"59.690625\" y=\"144.496796\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_7\">\n",
|
||
" <!-- 0 -->\n",
|
||
" <g transform=\"translate(46.328125 148.296015)scale(0.1 -0.1)\">\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"ytick_2\">\n",
|
||
" <g id=\"line2d_7\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#m7f47d3da76\" x=\"59.690625\" y=\"117.797281\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_8\">\n",
|
||
" <!-- 5000 -->\n",
|
||
" <g transform=\"translate(27.240625 121.5965)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-35\" d=\"M 691 4666 \n",
|
||
"L 3169 4666 \n",
|
||
"L 3169 4134 \n",
|
||
"L 1269 4134 \n",
|
||
"L 1269 2991 \n",
|
||
"Q 1406 3038 1543 3061 \n",
|
||
"Q 1681 3084 1819 3084 \n",
|
||
"Q 2600 3084 3056 2656 \n",
|
||
"Q 3513 2228 3513 1497 \n",
|
||
"Q 3513 744 3044 326 \n",
|
||
"Q 2575 -91 1722 -91 \n",
|
||
"Q 1428 -91 1123 -41 \n",
|
||
"Q 819 9 494 109 \n",
|
||
"L 494 744 \n",
|
||
"Q 775 591 1075 516 \n",
|
||
"Q 1375 441 1709 441 \n",
|
||
"Q 2250 441 2565 725 \n",
|
||
"Q 2881 1009 2881 1497 \n",
|
||
"Q 2881 1984 2565 2268 \n",
|
||
"Q 2250 2553 1709 2553 \n",
|
||
"Q 1456 2553 1204 2497 \n",
|
||
"Q 953 2441 691 2322 \n",
|
||
"L 691 4666 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-35\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"63.623047\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"127.246094\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"190.869141\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"ytick_3\">\n",
|
||
" <g id=\"line2d_8\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#m7f47d3da76\" x=\"59.690625\" y=\"91.097765\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_9\">\n",
|
||
" <!-- 10000 -->\n",
|
||
" <g transform=\"translate(20.878125 94.896984)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-31\" d=\"M 794 531 \n",
|
||
"L 1825 531 \n",
|
||
"L 1825 4091 \n",
|
||
"L 703 3866 \n",
|
||
"L 703 4441 \n",
|
||
"L 1819 4666 \n",
|
||
"L 2450 4666 \n",
|
||
"L 2450 531 \n",
|
||
"L 3481 531 \n",
|
||
"L 3481 0 \n",
|
||
"L 794 0 \n",
|
||
"L 794 531 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-31\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"63.623047\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"127.246094\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"190.869141\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"254.492188\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"ytick_4\">\n",
|
||
" <g id=\"line2d_9\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#m7f47d3da76\" x=\"59.690625\" y=\"64.39825\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_10\">\n",
|
||
" <!-- 15000 -->\n",
|
||
" <g transform=\"translate(20.878125 68.197469)scale(0.1 -0.1)\">\n",
|
||
" <use xlink:href=\"#DejaVuSans-31\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-35\" x=\"63.623047\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"127.246094\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"190.869141\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"254.492188\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"ytick_5\">\n",
|
||
" <g id=\"line2d_10\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#m7f47d3da76\" x=\"59.690625\" y=\"37.698734\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_11\">\n",
|
||
" <!-- 20000 -->\n",
|
||
" <g transform=\"translate(20.878125 41.497953)scale(0.1 -0.1)\">\n",
|
||
" <use xlink:href=\"#DejaVuSans-32\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"63.623047\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"127.246094\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"190.869141\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"254.492188\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"ytick_6\">\n",
|
||
" <g id=\"line2d_11\">\n",
|
||
" <g>\n",
|
||
" <use xlink:href=\"#m7f47d3da76\" x=\"59.690625\" y=\"10.999219\" style=\"stroke: #000000; stroke-width: 0.8\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_12\">\n",
|
||
" <!-- 25000 -->\n",
|
||
" <g transform=\"translate(20.878125 14.798437)scale(0.1 -0.1)\">\n",
|
||
" <use xlink:href=\"#DejaVuSans-32\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-35\" x=\"63.623047\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"127.246094\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"190.869141\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-30\" x=\"254.492188\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_13\">\n",
|
||
" <!-- count -->\n",
|
||
" <g transform=\"translate(14.798438 90.653046)rotate(-90)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-75\" d=\"M 544 1381 \n",
|
||
"L 544 3500 \n",
|
||
"L 1119 3500 \n",
|
||
"L 1119 1403 \n",
|
||
"Q 1119 906 1312 657 \n",
|
||
"Q 1506 409 1894 409 \n",
|
||
"Q 2359 409 2629 706 \n",
|
||
"Q 2900 1003 2900 1516 \n",
|
||
"L 2900 3500 \n",
|
||
"L 3475 3500 \n",
|
||
"L 3475 0 \n",
|
||
"L 2900 0 \n",
|
||
"L 2900 538 \n",
|
||
"Q 2691 219 2414 64 \n",
|
||
"Q 2138 -91 1772 -91 \n",
|
||
"Q 1169 -91 856 284 \n",
|
||
"Q 544 659 544 1381 \n",
|
||
"z\n",
|
||
"M 1991 3584 \n",
|
||
"L 1991 3584 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-63\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6f\" x=\"54.980469\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-75\" x=\"116.162109\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6e\" x=\"179.541016\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-74\" x=\"242.919922\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_23\">\n",
|
||
" <path d=\"M 59.690625 144.496796 \n",
|
||
"L 59.690625 8.596796 \n",
|
||
"\" style=\"fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_24\">\n",
|
||
" <path d=\"M 254.990625 144.496796 \n",
|
||
"L 254.990625 8.596796 \n",
|
||
"\" style=\"fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_25\">\n",
|
||
" <path d=\"M 59.690625 144.496796 \n",
|
||
"L 254.990625 144.496796 \n",
|
||
"\" style=\"fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_26\">\n",
|
||
" <path d=\"M 59.690625 8.596796 \n",
|
||
"L 254.990625 8.596796 \n",
|
||
"\" style=\"fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"legend_1\">\n",
|
||
" <g id=\"patch_27\">\n",
|
||
" <path d=\"M 155.389063 45.953046 \n",
|
||
"L 247.990625 45.953046 \n",
|
||
"Q 249.990625 45.953046 249.990625 43.953046 \n",
|
||
"L 249.990625 15.596796 \n",
|
||
"Q 249.990625 13.596796 247.990625 13.596796 \n",
|
||
"L 155.389063 13.596796 \n",
|
||
"Q 153.389063 13.596796 153.389063 15.596796 \n",
|
||
"L 153.389063 43.953046 \n",
|
||
"Q 153.389063 45.953046 155.389063 45.953046 \n",
|
||
"z\n",
|
||
"\" style=\"fill: #ffffff; opacity: 0.8; stroke: #cccccc; stroke-linejoin: miter\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_28\">\n",
|
||
" <path d=\"M 157.389063 25.195234 \n",
|
||
"L 177.389063 25.195234 \n",
|
||
"L 177.389063 18.195234 \n",
|
||
"L 157.389063 18.195234 \n",
|
||
"z\n",
|
||
"\" style=\"fill: #1f77b4\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_14\">\n",
|
||
" <!-- origin -->\n",
|
||
" <g transform=\"translate(185.389063 25.195234)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-69\" d=\"M 603 3500 \n",
|
||
"L 1178 3500 \n",
|
||
"L 1178 0 \n",
|
||
"L 603 0 \n",
|
||
"L 603 3500 \n",
|
||
"z\n",
|
||
"M 603 4863 \n",
|
||
"L 1178 4863 \n",
|
||
"L 1178 4134 \n",
|
||
"L 603 4134 \n",
|
||
"L 603 4863 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-67\" d=\"M 2906 1791 \n",
|
||
"Q 2906 2416 2648 2759 \n",
|
||
"Q 2391 3103 1925 3103 \n",
|
||
"Q 1463 3103 1205 2759 \n",
|
||
"Q 947 2416 947 1791 \n",
|
||
"Q 947 1169 1205 825 \n",
|
||
"Q 1463 481 1925 481 \n",
|
||
"Q 2391 481 2648 825 \n",
|
||
"Q 2906 1169 2906 1791 \n",
|
||
"z\n",
|
||
"M 3481 434 \n",
|
||
"Q 3481 -459 3084 -895 \n",
|
||
"Q 2688 -1331 1869 -1331 \n",
|
||
"Q 1566 -1331 1297 -1286 \n",
|
||
"Q 1028 -1241 775 -1147 \n",
|
||
"L 775 -588 \n",
|
||
"Q 1028 -725 1275 -790 \n",
|
||
"Q 1522 -856 1778 -856 \n",
|
||
"Q 2344 -856 2625 -561 \n",
|
||
"Q 2906 -266 2906 331 \n",
|
||
"L 2906 616 \n",
|
||
"Q 2728 306 2450 153 \n",
|
||
"Q 2172 0 1784 0 \n",
|
||
"Q 1141 0 747 490 \n",
|
||
"Q 353 981 353 1791 \n",
|
||
"Q 353 2603 747 3093 \n",
|
||
"Q 1141 3584 1784 3584 \n",
|
||
"Q 2172 3584 2450 3431 \n",
|
||
"Q 2728 3278 2906 2969 \n",
|
||
"L 2906 3500 \n",
|
||
"L 3481 3500 \n",
|
||
"L 3481 434 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6f\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-72\" x=\"61.181641\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-69\" x=\"102.294922\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-67\" x=\"130.078125\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-69\" x=\"193.554688\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6e\" x=\"221.337891\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <g id=\"patch_29\">\n",
|
||
" <path d=\"M 157.389063 39.873359 \n",
|
||
"L 177.389063 39.873359 \n",
|
||
"L 177.389063 32.873359 \n",
|
||
"L 157.389063 32.873359 \n",
|
||
"z\n",
|
||
"\" style=\"fill: url(#hdd73a5dc4c)\"/>\n",
|
||
" </g>\n",
|
||
" <g id=\"text_15\">\n",
|
||
" <!-- subsampled -->\n",
|
||
" <g transform=\"translate(185.389063 39.873359)scale(0.1 -0.1)\">\n",
|
||
" <defs>\n",
|
||
" <path id=\"DejaVuSans-62\" d=\"M 3116 1747 \n",
|
||
"Q 3116 2381 2855 2742 \n",
|
||
"Q 2594 3103 2138 3103 \n",
|
||
"Q 1681 3103 1420 2742 \n",
|
||
"Q 1159 2381 1159 1747 \n",
|
||
"Q 1159 1113 1420 752 \n",
|
||
"Q 1681 391 2138 391 \n",
|
||
"Q 2594 391 2855 752 \n",
|
||
"Q 3116 1113 3116 1747 \n",
|
||
"z\n",
|
||
"M 1159 2969 \n",
|
||
"Q 1341 3281 1617 3432 \n",
|
||
"Q 1894 3584 2278 3584 \n",
|
||
"Q 2916 3584 3314 3078 \n",
|
||
"Q 3713 2572 3713 1747 \n",
|
||
"Q 3713 922 3314 415 \n",
|
||
"Q 2916 -91 2278 -91 \n",
|
||
"Q 1894 -91 1617 61 \n",
|
||
"Q 1341 213 1159 525 \n",
|
||
"L 1159 0 \n",
|
||
"L 581 0 \n",
|
||
"L 581 4863 \n",
|
||
"L 1159 4863 \n",
|
||
"L 1159 2969 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-61\" d=\"M 2194 1759 \n",
|
||
"Q 1497 1759 1228 1600 \n",
|
||
"Q 959 1441 959 1056 \n",
|
||
"Q 959 750 1161 570 \n",
|
||
"Q 1363 391 1709 391 \n",
|
||
"Q 2188 391 2477 730 \n",
|
||
"Q 2766 1069 2766 1631 \n",
|
||
"L 2766 1759 \n",
|
||
"L 2194 1759 \n",
|
||
"z\n",
|
||
"M 3341 1997 \n",
|
||
"L 3341 0 \n",
|
||
"L 2766 0 \n",
|
||
"L 2766 531 \n",
|
||
"Q 2569 213 2275 61 \n",
|
||
"Q 1981 -91 1556 -91 \n",
|
||
"Q 1019 -91 701 211 \n",
|
||
"Q 384 513 384 1019 \n",
|
||
"Q 384 1609 779 1909 \n",
|
||
"Q 1175 2209 1959 2209 \n",
|
||
"L 2766 2209 \n",
|
||
"L 2766 2266 \n",
|
||
"Q 2766 2663 2505 2880 \n",
|
||
"Q 2244 3097 1772 3097 \n",
|
||
"Q 1472 3097 1187 3025 \n",
|
||
"Q 903 2953 641 2809 \n",
|
||
"L 641 3341 \n",
|
||
"Q 956 3463 1253 3523 \n",
|
||
"Q 1550 3584 1831 3584 \n",
|
||
"Q 2591 3584 2966 3190 \n",
|
||
"Q 3341 2797 3341 1997 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-6d\" d=\"M 3328 2828 \n",
|
||
"Q 3544 3216 3844 3400 \n",
|
||
"Q 4144 3584 4550 3584 \n",
|
||
"Q 5097 3584 5394 3201 \n",
|
||
"Q 5691 2819 5691 2113 \n",
|
||
"L 5691 0 \n",
|
||
"L 5113 0 \n",
|
||
"L 5113 2094 \n",
|
||
"Q 5113 2597 4934 2840 \n",
|
||
"Q 4756 3084 4391 3084 \n",
|
||
"Q 3944 3084 3684 2787 \n",
|
||
"Q 3425 2491 3425 1978 \n",
|
||
"L 3425 0 \n",
|
||
"L 2847 0 \n",
|
||
"L 2847 2094 \n",
|
||
"Q 2847 2600 2669 2842 \n",
|
||
"Q 2491 3084 2119 3084 \n",
|
||
"Q 1678 3084 1418 2786 \n",
|
||
"Q 1159 2488 1159 1978 \n",
|
||
"L 1159 0 \n",
|
||
"L 581 0 \n",
|
||
"L 581 3500 \n",
|
||
"L 1159 3500 \n",
|
||
"L 1159 2956 \n",
|
||
"Q 1356 3278 1631 3431 \n",
|
||
"Q 1906 3584 2284 3584 \n",
|
||
"Q 2666 3584 2933 3390 \n",
|
||
"Q 3200 3197 3328 2828 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-6c\" d=\"M 603 4863 \n",
|
||
"L 1178 4863 \n",
|
||
"L 1178 0 \n",
|
||
"L 603 0 \n",
|
||
"L 603 4863 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" <path id=\"DejaVuSans-64\" d=\"M 2906 2969 \n",
|
||
"L 2906 4863 \n",
|
||
"L 3481 4863 \n",
|
||
"L 3481 0 \n",
|
||
"L 2906 0 \n",
|
||
"L 2906 525 \n",
|
||
"Q 2725 213 2448 61 \n",
|
||
"Q 2172 -91 1784 -91 \n",
|
||
"Q 1150 -91 751 415 \n",
|
||
"Q 353 922 353 1747 \n",
|
||
"Q 353 2572 751 3078 \n",
|
||
"Q 1150 3584 1784 3584 \n",
|
||
"Q 2172 3584 2448 3432 \n",
|
||
"Q 2725 3281 2906 2969 \n",
|
||
"z\n",
|
||
"M 947 1747 \n",
|
||
"Q 947 1113 1208 752 \n",
|
||
"Q 1469 391 1925 391 \n",
|
||
"Q 2381 391 2643 752 \n",
|
||
"Q 2906 1113 2906 1747 \n",
|
||
"Q 2906 2381 2643 2742 \n",
|
||
"Q 2381 3103 1925 3103 \n",
|
||
"Q 1469 3103 1208 2742 \n",
|
||
"Q 947 2381 947 1747 \n",
|
||
"z\n",
|
||
"\" transform=\"scale(0.015625)\"/>\n",
|
||
" </defs>\n",
|
||
" <use xlink:href=\"#DejaVuSans-73\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-75\" x=\"52.099609\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-62\" x=\"115.478516\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-73\" x=\"178.955078\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-61\" x=\"231.054688\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6d\" x=\"292.333984\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-70\" x=\"389.746094\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-6c\" x=\"453.222656\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-65\" x=\"481.005859\"/>\n",
|
||
" <use xlink:href=\"#DejaVuSans-64\" x=\"542.529297\"/>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" </g>\n",
|
||
" <defs>\n",
|
||
" <clipPath id=\"p224370b0b9\">\n",
|
||
" <rect x=\"59.690625\" y=\"8.596796\" width=\"195.3\" height=\"135.9\"/>\n",
|
||
" </clipPath>\n",
|
||
" </defs>\n",
|
||
" <defs>\n",
|
||
" <pattern id=\"hdd73a5dc4c\" patternUnits=\"userSpaceOnUse\" x=\"0\" y=\"0\" width=\"72\" height=\"72\">\n",
|
||
" <rect x=\"0\" y=\"0\" width=\"73\" height=\"73\" fill=\"#ff7f0e\"/>\n",
|
||
" <path d=\"M -36 36 \n",
|
||
"L 36 -36 \n",
|
||
"M -24 48 \n",
|
||
"L 48 -24 \n",
|
||
"M -12 60 \n",
|
||
"L 60 -12 \n",
|
||
"M 0 72 \n",
|
||
"L 72 0 \n",
|
||
"M 12 84 \n",
|
||
"L 84 12 \n",
|
||
"M 24 96 \n",
|
||
"L 96 24 \n",
|
||
"M 36 108 \n",
|
||
"L 108 36 \n",
|
||
"\" style=\"fill: #000000; stroke: #000000; stroke-width: 1.0; stroke-linecap: butt; stroke-linejoin: miter\"/>\n",
|
||
" </pattern>\n",
|
||
" </defs>\n",
|
||
"</svg>\n"
|
||
],
|
||
"text/plain": [
|
||
"<Figure size 252x180 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {
|
||
"needs_background": "light"
|
||
},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"d2l.show_list_len_pair_hist(\n",
|
||
" ['origin', 'subsampled'], '# tokens per sentence',\n",
|
||
" 'count', sentences, subsampled);"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "80da6e9d",
|
||
"metadata": {
|
||
"origin_pos": 12
|
||
},
|
||
"source": [
|
||
"对于单个词元，高频词“the”的采样率不到1/20。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 6,
|
||
"id": "2ac63b1a",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:44.277942Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:44.277661Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:44.319135Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:44.317982Z"
|
||
},
|
||
"origin_pos": 13,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"'\"the\"的数量：之前=50770, 之后=2056'"
|
||
]
|
||
},
|
||
"execution_count": 6,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"def compare_counts(token):\n",
|
||
" return (f'\"{token}\"的数量：'\n",
|
||
" f'之前={sum([l.count(token) for l in sentences])}, '\n",
|
||
" f'之后={sum([l.count(token) for l in subsampled])}')\n",
|
||
"\n",
|
||
"compare_counts('the')"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "73ef69ef",
|
||
"metadata": {
|
||
"origin_pos": 14
|
||
},
|
||
"source": [
|
||
"相比之下，低频词“join”则被完全保留。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 7,
|
||
"id": "9307cb04",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:44.324650Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:44.323726Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:44.366586Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:44.365449Z"
|
||
},
|
||
"origin_pos": 15,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"'\"join\"的数量：之前=45, 之后=45'"
|
||
]
|
||
},
|
||
"execution_count": 7,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"compare_counts('join')"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "38762200",
|
||
"metadata": {
|
||
"origin_pos": 16
|
||
},
|
||
"source": [
|
||
"在下采样之后，我们将词元映射到它们在语料库中的索引。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 8,
|
||
"id": "ed59e4d0",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:44.371681Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:44.370695Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:44.930927Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:44.929824Z"
|
||
},
|
||
"origin_pos": 17,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"[[], [2115, 274, 406], [140, 3, 5277, 3054, 1580]]"
|
||
]
|
||
},
|
||
"execution_count": 8,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"corpus = [vocab[line] for line in subsampled]\n",
|
||
"corpus[:3]"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "fe5918fb",
|
||
"metadata": {
|
||
"origin_pos": 18
|
||
},
|
||
"source": [
|
||
"## 中心词和上下文词的提取\n",
|
||
"\n",
|
||
"下面的`get_centers_and_contexts`函数从`corpus`中提取所有中心词及其上下文词。它随机采样1到`max_window_size`之间的整数作为上下文窗口。对于任一中心词，与其距离不超过采样上下文窗口大小的词为其上下文词。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 9,
|
||
"id": "d4a20ba3",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:44.935833Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:44.935066Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:44.944963Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:44.943901Z"
|
||
},
|
||
"origin_pos": 19,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"#@save\n",
|
||
"def get_centers_and_contexts(corpus, max_window_size):\n",
|
||
" \"\"\"返回跳元模型中的中心词和上下文词\"\"\"\n",
|
||
" centers, contexts = [], []\n",
|
||
" for line in corpus:\n",
|
||
" # 要形成“中心词-上下文词”对，每个句子至少需要有2个词\n",
|
||
" if len(line) < 2:\n",
|
||
" continue\n",
|
||
" centers += line\n",
|
||
" for i in range(len(line)): # 上下文窗口中间i\n",
|
||
" window_size = random.randint(1, max_window_size)\n",
|
||
" indices = list(range(max(0, i - window_size),\n",
|
||
" min(len(line), i + 1 + window_size)))\n",
|
||
" # 从上下文词中排除中心词\n",
|
||
" indices.remove(i)\n",
|
||
" contexts.append([line[idx] for idx in indices])\n",
|
||
" return centers, contexts"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "86fba895",
|
||
"metadata": {
|
||
"origin_pos": 20
|
||
},
|
||
"source": [
|
||
"接下来，我们创建一个人工数据集，分别包含7个和3个单词的两个句子。设置最大上下文窗口大小为2，并打印所有中心词及其上下文词。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 10,
|
||
"id": "fae4771b",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:44.948910Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:44.948190Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:44.955563Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:44.954488Z"
|
||
},
|
||
"origin_pos": 21,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"数据集 [[0, 1, 2, 3, 4, 5, 6], [7, 8, 9]]\n",
|
||
"中心词 0 的上下文词是 [1]\n",
|
||
"中心词 1 的上下文词是 [0, 2]\n",
|
||
"中心词 2 的上下文词是 [0, 1, 3, 4]\n",
|
||
"中心词 3 的上下文词是 [2, 4]\n",
|
||
"中心词 4 的上下文词是 [3, 5]\n",
|
||
"中心词 5 的上下文词是 [4, 6]\n",
|
||
"中心词 6 的上下文词是 [5]\n",
|
||
"中心词 7 的上下文词是 [8, 9]\n",
|
||
"中心词 8 的上下文词是 [7, 9]\n",
|
||
"中心词 9 的上下文词是 [7, 8]\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"tiny_dataset = [list(range(7)), list(range(7, 10))]\n",
|
||
"print('数据集', tiny_dataset)\n",
|
||
"for center, context in zip(*get_centers_and_contexts(tiny_dataset, 2)):\n",
|
||
" print('中心词', center, '的上下文词是', context)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "e21272fc",
|
||
"metadata": {
|
||
"origin_pos": 22
|
||
},
|
||
"source": [
|
||
"在PTB数据集上进行训练时，我们将最大上下文窗口大小设置为5。下面提取数据集中的所有中心词及其上下文词。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 11,
|
||
"id": "ec92f27e",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:44.960145Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:44.959231Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:47.218796Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:47.217626Z"
|
||
},
|
||
"origin_pos": 23,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"'# “中心词-上下文词对”的数量: 1499984'"
|
||
]
|
||
},
|
||
"execution_count": 11,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"all_centers, all_contexts = get_centers_and_contexts(corpus, 5)\n",
|
||
"f'# “中心词-上下文词对”的数量: {sum([len(contexts) for contexts in all_contexts])}'"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "f48c535f",
|
||
"metadata": {
|
||
"origin_pos": 24
|
||
},
|
||
"source": [
|
||
"## 负采样\n",
|
||
"\n",
|
||
"我们使用负采样进行近似训练。为了根据预定义的分布对噪声词进行采样，我们定义以下`RandomGenerator`类，其中（可能未规范化的）采样分布通过变量`sampling_weights`传递。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 12,
|
||
"id": "365189a2",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:47.223801Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:47.223354Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:47.232254Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:47.231166Z"
|
||
},
|
||
"origin_pos": 25,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"#@save\n",
|
||
"class RandomGenerator:\n",
|
||
" \"\"\"根据n个采样权重在{1,...,n}中随机抽取\"\"\"\n",
|
||
" def __init__(self, sampling_weights):\n",
|
||
" # Exclude\n",
|
||
" self.population = list(range(1, len(sampling_weights) + 1))\n",
|
||
" self.sampling_weights = sampling_weights\n",
|
||
" self.candidates = []\n",
|
||
" self.i = 0\n",
|
||
"\n",
|
||
" def draw(self):\n",
|
||
" if self.i == len(self.candidates):\n",
|
||
" # 缓存k个随机采样结果\n",
|
||
" self.candidates = random.choices(\n",
|
||
" self.population, self.sampling_weights, k=10000)\n",
|
||
" self.i = 0\n",
|
||
" self.i += 1\n",
|
||
" return self.candidates[self.i - 1]"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "f886ada9",
|
||
"metadata": {
|
||
"origin_pos": 26
|
||
},
|
||
"source": [
|
||
"例如，我们可以在索引1、2和3中绘制10个随机变量$X$，采样概率为$P(X=1)=2/9, P(X=2)=3/9$和$P(X=3)=4/9$，如下所示。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 13,
|
||
"id": "f534865c",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:47.237153Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:47.236381Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:47.251510Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:47.250435Z"
|
||
},
|
||
"origin_pos": 27,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"[1, 2, 2, 3, 3, 3, 3, 2, 1, 2]"
|
||
]
|
||
},
|
||
"execution_count": 13,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"#@save\n",
|
||
"generator = RandomGenerator([2, 3, 4])\n",
|
||
"[generator.draw() for _ in range(10)]"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "fe4049d4",
|
||
"metadata": {
|
||
"origin_pos": 28
|
||
},
|
||
"source": [
|
||
"对于一对中心词和上下文词，我们随机抽取了`K`个（实验中为5个）噪声词。根据word2vec论文中的建议，将噪声词$w$的采样概率$P(w)$设置为其在字典中的相对频率，其幂为0.75 :cite:`Mikolov.Sutskever.Chen.ea.2013`。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 14,
|
||
"id": "21950025",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:47.256344Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:47.255586Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:59.259799Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:59.258793Z"
|
||
},
|
||
"origin_pos": 29,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"#@save\n",
|
||
"def get_negatives(all_contexts, vocab, counter, K):\n",
|
||
" \"\"\"返回负采样中的噪声词\"\"\"\n",
|
||
" # 索引为1、2、...（索引0是词表中排除的未知标记）\n",
|
||
" sampling_weights = [counter[vocab.to_tokens(i)]**0.75\n",
|
||
" for i in range(1, len(vocab))]\n",
|
||
" all_negatives, generator = [], RandomGenerator(sampling_weights)\n",
|
||
" for contexts in all_contexts:\n",
|
||
" negatives = []\n",
|
||
" while len(negatives) < len(contexts) * K:\n",
|
||
" neg = generator.draw()\n",
|
||
" # 噪声词不能是上下文词\n",
|
||
" if neg not in contexts:\n",
|
||
" negatives.append(neg)\n",
|
||
" all_negatives.append(negatives)\n",
|
||
" return all_negatives\n",
|
||
"\n",
|
||
"all_negatives = get_negatives(all_contexts, vocab, counter, 5)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "8aa17e2d",
|
||
"metadata": {
|
||
"origin_pos": 30
|
||
},
|
||
"source": [
|
||
"## 小批量加载训练实例\n",
|
||
":label:`subsec_word2vec-minibatch-loading`\n",
|
||
"\n",
|
||
"在提取所有中心词及其上下文词和采样噪声词后，将它们转换成小批量的样本，在训练过程中可以迭代加载。\n",
|
||
"\n",
|
||
"在小批量中，$i^\\mathrm{th}$个样本包括中心词及其$n_i$个上下文词和$m_i$个噪声词。由于上下文窗口大小不同，$n_i+m_i$对于不同的$i$是不同的。因此，对于每个样本，我们在`contexts_negatives`个变量中将其上下文词和噪声词连结起来，并填充零，直到连结长度达到$\\max_i n_i+m_i$(`max_len`)。为了在计算损失时排除填充，我们定义了掩码变量`masks`。在`masks`中的元素和`contexts_negatives`中的元素之间存在一一对应关系，其中`masks`中的0（否则为1）对应于`contexts_negatives`中的填充。\n",
|
||
"\n",
|
||
"为了区分正反例，我们在`contexts_negatives`中通过一个`labels`变量将上下文词与噪声词分开。类似于`masks`，在`labels`中的元素和`contexts_negatives`中的元素之间也存在一一对应关系，其中`labels`中的1（否则为0）对应于`contexts_negatives`中的上下文词的正例。\n",
|
||
"\n",
|
||
"上述思想在下面的`batchify`函数中实现。其输入`data`是长度等于批量大小的列表，其中每个元素是由中心词`center`、其上下文词`context`和其噪声词`negative`组成的样本。此函数返回一个可以在训练期间加载用于计算的小批量，例如包括掩码变量。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 15,
|
||
"id": "8e92a65e",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:59.264970Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:59.264337Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:59.271417Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:59.270518Z"
|
||
},
|
||
"origin_pos": 31,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"#@save\n",
|
||
"def batchify(data):\n",
|
||
" \"\"\"返回带有负采样的跳元模型的小批量样本\"\"\"\n",
|
||
" max_len = max(len(c) + len(n) for _, c, n in data)\n",
|
||
" centers, contexts_negatives, masks, labels = [], [], [], []\n",
|
||
" for center, context, negative in data:\n",
|
||
" cur_len = len(context) + len(negative)\n",
|
||
" centers += [center]\n",
|
||
" contexts_negatives += \\\n",
|
||
" [context + negative + [0] * (max_len - cur_len)]\n",
|
||
" masks += [[1] * cur_len + [0] * (max_len - cur_len)]\n",
|
||
" labels += [[1] * len(context) + [0] * (max_len - len(context))]\n",
|
||
" return (torch.tensor(centers).reshape((-1, 1)), torch.tensor(\n",
|
||
" contexts_negatives), torch.tensor(masks), torch.tensor(labels))"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "7aeb5c51",
|
||
"metadata": {
|
||
"origin_pos": 32
|
||
},
|
||
"source": [
|
||
"让我们使用一个小批量的两个样本来测试此函数。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 16,
|
||
"id": "e14e34ce",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:59.276193Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:59.275387Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:59.282832Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:59.281912Z"
|
||
},
|
||
"origin_pos": 33,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"centers = tensor([[1],\n",
|
||
" [1]])\n",
|
||
"contexts_negatives = tensor([[2, 2, 3, 3, 3, 3],\n",
|
||
" [2, 2, 2, 3, 3, 0]])\n",
|
||
"masks = tensor([[1, 1, 1, 1, 1, 1],\n",
|
||
" [1, 1, 1, 1, 1, 0]])\n",
|
||
"labels = tensor([[1, 1, 0, 0, 0, 0],\n",
|
||
" [1, 1, 1, 0, 0, 0]])\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"x_1 = (1, [2, 2], [3, 3, 3, 3])\n",
|
||
"x_2 = (1, [2, 2, 2], [3, 3])\n",
|
||
"batch = batchify((x_1, x_2))\n",
|
||
"\n",
|
||
"names = ['centers', 'contexts_negatives', 'masks', 'labels']\n",
|
||
"for name, data in zip(names, batch):\n",
|
||
" print(name, '=', data)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "e1eef3d8",
|
||
"metadata": {
|
||
"origin_pos": 34
|
||
},
|
||
"source": [
|
||
"## 整合代码\n",
|
||
"\n",
|
||
"最后，我们定义了读取PTB数据集并返回数据迭代器和词表的`load_data_ptb`函数。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 17,
|
||
"id": "8ddfb20d",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:59.287587Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:59.286823Z",
|
||
"iopub.status.idle": "2023-08-18T07:01:59.296040Z",
|
||
"shell.execute_reply": "2023-08-18T07:01:59.294978Z"
|
||
},
|
||
"origin_pos": 36,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"#@save\n",
|
||
"def load_data_ptb(batch_size, max_window_size, num_noise_words):\n",
|
||
" \"\"\"下载PTB数据集，然后将其加载到内存中\"\"\"\n",
|
||
" num_workers = d2l.get_dataloader_workers()\n",
|
||
" sentences = read_ptb()\n",
|
||
" vocab = d2l.Vocab(sentences, min_freq=10)\n",
|
||
" subsampled, counter = subsample(sentences, vocab)\n",
|
||
" corpus = [vocab[line] for line in subsampled]\n",
|
||
" all_centers, all_contexts = get_centers_and_contexts(\n",
|
||
" corpus, max_window_size)\n",
|
||
" all_negatives = get_negatives(\n",
|
||
" all_contexts, vocab, counter, num_noise_words)\n",
|
||
"\n",
|
||
" class PTBDataset(torch.utils.data.Dataset):\n",
|
||
" def __init__(self, centers, contexts, negatives):\n",
|
||
" assert len(centers) == len(contexts) == len(negatives)\n",
|
||
" self.centers = centers\n",
|
||
" self.contexts = contexts\n",
|
||
" self.negatives = negatives\n",
|
||
"\n",
|
||
" def __getitem__(self, index):\n",
|
||
" return (self.centers[index], self.contexts[index],\n",
|
||
" self.negatives[index])\n",
|
||
"\n",
|
||
" def __len__(self):\n",
|
||
" return len(self.centers)\n",
|
||
"\n",
|
||
" dataset = PTBDataset(all_centers, all_contexts, all_negatives)\n",
|
||
"\n",
|
||
" data_iter = torch.utils.data.DataLoader(\n",
|
||
" dataset, batch_size, shuffle=True,\n",
|
||
" collate_fn=batchify, num_workers=num_workers)\n",
|
||
" return data_iter, vocab"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "97991d10",
|
||
"metadata": {
|
||
"origin_pos": 38
|
||
},
|
||
"source": [
|
||
"让我们打印数据迭代器的第一个小批量。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 18,
|
||
"id": "5115b257",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2023-08-18T07:01:59.300574Z",
|
||
"iopub.status.busy": "2023-08-18T07:01:59.299960Z",
|
||
"iopub.status.idle": "2023-08-18T07:02:13.672095Z",
|
||
"shell.execute_reply": "2023-08-18T07:02:13.671142Z"
|
||
},
|
||
"origin_pos": 39,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"centers shape: torch.Size([512, 1])\n",
|
||
"contexts_negatives shape: torch.Size([512, 60])\n",
|
||
"masks shape: torch.Size([512, 60])\n",
|
||
"labels shape: torch.Size([512, 60])\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"data_iter, vocab = load_data_ptb(512, 5, 5)\n",
|
||
"for batch in data_iter:\n",
|
||
" for name, data in zip(names, batch):\n",
|
||
" print(name, 'shape:', data.shape)\n",
|
||
" break"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "cfc03f54",
|
||
"metadata": {
|
||
"origin_pos": 40
|
||
},
|
||
"source": [
|
||
"## 小结\n",
|
||
"\n",
|
||
"* 高频词在训练中可能不是那么有用。我们可以对他们进行下采样，以便在训练中加快速度。\n",
|
||
"* 为了提高计算效率，我们以小批量方式加载样本。我们可以定义其他变量来区分填充标记和非填充标记，以及正例和负例。\n",
|
||
"\n",
|
||
"## 练习\n",
|
||
"\n",
|
||
"1. 如果不使用下采样，本节中代码的运行时间会发生什么变化？\n",
|
||
"1. `RandomGenerator`类缓存`k`个随机采样结果。将`k`设置为其他值，看看它如何影响数据加载速度。\n",
|
||
"1. 本节代码中的哪些其他超参数可能会影响数据加载速度？\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "9e415387",
|
||
"metadata": {
|
||
"origin_pos": 42,
|
||
"tab": [
|
||
"pytorch"
|
||
]
|
||
},
|
||
"source": [
|
||
"[Discussions](https://discuss.d2l.ai/t/5735)\n"
|
||
]
|
||
}
|
||
],
|
||
"metadata": {
|
||
"language_info": {
|
||
"name": "python"
|
||
},
|
||
"required_libs": []
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 5
|
||
} |