trope_examples_dataset.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "7fc9ae11-6f84-4ed6-987f-9e6ca6ad78ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "pd.set_option('display.max_colwidth', None)\n",
    "import numpy as np\n",
    "from tqdm.auto import tqdm\n",
    "tqdm.pandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c3d40ab1-392b-4633-b234-cc12e5bfbe43",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_path = \"dataset/my_tropes.csv\"\n",
    "tropes = pd.read_csv(dataset_path, dtype={\"TropeID\": str, \"Trope\": str, \"Description\": str})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7fcf3be6-aa60-43e8-9454-571fc5357042",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>TropeID</th>\n",
       "      <th>Trope</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>t00079</td>\n",
       "      <td>AbusiveParents</td>\n",
       "      <td>Parents are supposed to be the protectors of children, but these parents are either so damaged themselves that they can't do the job, greedy or villainous to the point that they never had any interest in doing the job properly, or would rather use the child as a means to an end. Sometimes they're just sadistic assholes.\\nThis includes parents who are emotionally, verbally, physically, or mentally abusive, or who neglectfully allow their children to be abused by others if they don't abuse the child themselves; sexual abuse is typically treated as a special kind of evil. Sometimes, the abuse at the hands of their parents becomes a Freudian Excuse for a villain. Other times, the character manages to not grow up broken, bitter, and hateful, and instead a different and better person than the upbringing would incline one to think. Troubling Unchildlike Behavior is often a tell-tale sign that things are not right at home.\\nAbusive Parents are commonplace in fairy tales and Classical Mythology which makes this trope Older Than Feudalism. Note that The Brothers Grimm, when they collected European fairy tales, were uncomfortable with the idea of Abusive Parents and so frequently changed the Abusive Parents in the traditional stories into abusive step parents.\\nSometimes, a parent will go as far as to kill the child in question, in which case this is Offing the Offspring. In other ways cases, the parent's abuse occasionally drives the offspring to snap, commit Revenge and finally kill them, thus becoming a Self-Made Orphan. Although other times, the parents end up as a\\nKarma Houdini. Calling the Old Man Out occurs when a fed-up child retaliates with a \"The Reason You Suck\" Speech. If the child gets out of the broken family and forms healthy friendships, but reacts badly when their abusive parents show up again, well, Friends Are Chosen, Family Aren't.\\nBear in mind that not everyone agrees on the line between actual abuse and merely heavy-handed parenting (or even normal parenting). Is Moving the Goalposts merely inspiring the child to achieve more, or the most insidious form of abuse to instill mistrust and paranoia to the children? Some include spanking as abuse; others think it's appropriate given certain guidelines. Some believe it's okay to make a kid go without a meal (they won't starve that easily); others disagree. Making a kid miss a friend's birthday sleepover — is that emotional abuse? Raising a kid without exposure to TV? Telling your daughter she's getting fat? A little friendly name-calling? There's a line here somewhere, but not everyone agrees on where it is.\\nIf a parent has just dumped the child, for whatever reason, that's Parental Abandonment; if they aren't paying attention, that's Parental Neglect. If the parents refuse to discipline their kids, they are Pushover Parents. Contrast Mama Bear or Papa Wolf (where others abuse the children and the parents abuse the abusers), and the more extreme variant of Knight Templar Parent, where the abusive parent is violently overprotective. Abusive Precursors can be considered this on a metaphorical level. See Hilariously Abusive Childhood for when this is cranked up to absurd levels and played for laughs. Black Comedy is often connected in the comedic aspect of it, and a Big, Screwed-Up Family may be involved if it is adult comedy. In keeping with the above note, some may call the show on it and say Dude, Not Funny!. See Evil Matriarch and Archnemesis Dad for characters who are beyond abusive and outright evil. For parents who are mostly abused by their children, see Pushover Parents. Abusive-type parents are mostly wanted criminals being chased by the law.\\nWhile they do not have to be the child's actual, technical parents to be part of this trope, it's pretty important that they are closely related and live together, like a Wicked Stepmother or an Evil Uncle taking care of the Parentally Deprived. After all, it's much more disgusting that somebody related to the child could bring themselves to hurt them, rather than a mere foster family.\\nThe polar opposite, of course, are Good Parents.\\nNOTE: Please do not use this trope for complaining about parents you don't like.\\nUnfortunately, Abusive Parents are Truth in Television and a very sensitive topic for many, so please no Real Life examples - it is sufficient to say they do exist and they are all too common.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>t00190</td>\n",
       "      <td>ActionGirl</td>\n",
       "      <td>An Action Girl is a female badass who is tough and kicks butt. Damsel in Distress? Not her. She's featured in far more than the Designated Girl Fight. She faces dangerous foes and deadly obstacles, and she wins.\\nFor the longest time in many cultures, Double Standards in both fiction and real life meant that when it came to action and fighting, guys definitely outnumbered girls. Men Act, Women Are was the rule of thumb, which led to the Action Girl being a subversion of what was acceptable. As society has marched on, this view has faded in some media, leading to stories where Action Girls become the norm rather than the exception, such as stories set in a World of Badass, and especially a World of Action Girls.\\nThe broad Action Girl concept can take many forms. Faux Action Girl is a case of presenting a character as this, only for them not to live up to the standards of this trope. Dark Action Girl is the villainous variety, and Affirmative Action Girl is a cast addition intended to balance out gender ratios that typically also falls under this trope. Less action-y versions include You Go, Girl! and Plucky Girl.\\nThis is somewhat of a Cyclic Trope, with Action Girls often having surges in popularity in the 1940s, 1970s, 1990s and 2010s, probably not coincidentally times when the women's movement increased in prominence.\\nSee below for even more subtropes and other variants.\\nNot to be confused with Action Girl, a comic series about an Action Girl. For advice on how to write these types of characters, see Write An Action Girl.\\n\\nGenres and worlds populated by Action Girls:</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>t00263</td>\n",
       "      <td>AdaptationDistillation</td>\n",
       "      <td>Some adaptations take a complex character or situation and greatly simplify it, removing elements the producer believed to be unnecessary. This effect is more common when adapting from a long-running series, especially if it hasn't had a singular vision over the years.\\nCompare Compressed Adaptation, which deals more along the lines of a direct adaptation and is about all the stuff that is cut out along the way to make this trope happen.\\nContrast with Pragmatic Adaptation: in a distillation, a complex story is simplified, without much substantive change. In a Pragmatic Adaptation, the story is changed with the shift in medium. Also, contrast Adaptation Expansion, which tacks on more content but retains the original source material. When a story element is removed but its effects aren't, that's Adaptation Explanation Extrication.\\nThis isn't about adaptations that are better than the original. This is about elements that are simplified or removed in order to make an adaptation shorter.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>t00265</td>\n",
       "      <td>AdaptationExpansion</td>\n",
       "      <td>This is the complete opposite of Compressed Adaptation. It occurs when a short, very simple tale is given a much bigger adaptation, typically as a feature-length film. To be brought up to feature length, the storyline will have to be padded with some new stuff — a lot of new stuff. Cue Alternative Character Interpretations that require elaborate backstories, minor characters given much larger parts, completely new characters, and sometimes Plot Holes, Plot Tumors, and a triple dozen subplots that were not in the original work.\\nThis has a tendency to make the story unrecognizable as a retelling of the original. In some instances, the original story will end up as one small part of a much larger, more convoluted story. This will usually be the climax, in which case the film essentially gave you an hour or more of Back Story. This most often happens with movies based on novellas, short stories, video games or children's books. In the children's books instance, this can lead to the introduction of Darker and Edgier into a normally benign story or the reintroduction of elements lost to Disneyfication of classic stories.\\nWhile this trope is often associated with the fan complaint \"They Changed It, Now It Sucks!!\", it can be done well. Remember that Tropes Are Tools and extensions for adaptations might need to happen to fill the required runtime, just going about it right is the problem most productions face.\\nCompare Expanded Universe. See also Adaptation Distillation, Compressed Adaptation, Humble Beginnings, Patchwork Story, Not His Sled, Updated Re Release. An adaptation In Name Only goes even further than this, throwing out the original plot and making things up out of whole cloth.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>t00270</td>\n",
       "      <td>AdaptationNameChange</td>\n",
       "      <td>Alice Andrews in the books becomes Alice Allen in The Movie then Annie Adams in The Series.\\nThis is when a character or group has a name change between adaptations.\\nThe Name Change of The Character can be their first name, last name or both that also includes super alter egos.\\nThe key to this trope is that between adaptations, the character must clearly be the same. If they are a Suspiciously Similar Substitute or a Composite Character, then this trope does not apply.\\nThis can sometimes be an Enforced Trope. Production studios retain a team of lawyers whose job it is to ensure that every name in every work is shared either by many people or by no people in the area that the work is being produced, a process known as \"clearance\". This is to avoid a lawsuit: if there's exactly one person named Darkness von Gothick, you're just begging for him to sue you for portraying him as a villain. But names that are common in one place and time might be more rare in another, resulting in a name being cleared for use when The Movie was filmed in Hollywood fifty years ago but not when The Series is filmed in Vancouver today.\\nThis is also very common in foreign adaptations, where names are changed to either reflect the local culture, or to give the show its own identity. It may also accompany a Race Lift, Adaptational Nationality change, or Gender Flip, with a character's name reflecting his or her new ethnicity or gender (usually).\\nThis trope only applies when the adaptation is in the same language as the original. When a character's name changes in a translation, that's Dub Name Change. If a character didn't originally have any name then they're Named by the Adaptation. If it's Based on a True Story, see Roman à Clef.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>495</th>\n",
       "      <td>t30000</td>\n",
       "      <td>Redshirts</td>\n",
       "      <td>This is the Good Counterpart of Evil Minions and Mooks — set filler for our heroes' side. Their purpose is almost exclusively to give the writers someone to kill who isn't a main character, although they can also serve as Spear Carriers. In a series where The Main Characters Do Everything, if you suddenly see someone else who you've never seen before involved in the main story, they are probably Redshirts.\\nThey are used to show how the monster works, and demonstrate that it is indeed a deadly menace, without having to lose anyone important. Expect someone to say \"He's Dead, Jim\", lament this \"valued crew member's senseless death\", and then promptly forget him. Security personnel in general fall victim to the worst shade of this trope, as most of the time their deaths aren't even acknowledged at all; according to Hollywood, you could walk into a bank and shoot a security guard right in the face without anyone making a fuss. If you shot anyone else afterward, the headline would just read \"Bank Customers Killed\".\\nPlease note: this Trope is actually very inaccurate when you compare it to Real Life. If you were to watch every episode of Star Trek: The Original Series, count the number of casualties that the Enterprise had, and then compare that to an actual military, you'd see that Kirk's record as a leader in this regard is excellent, far better than any general in U.S. history. Even war heroes like George Washington and Dwight D. Eisenhower had proportionately more casualties among their troops.note It has been suggested that it would be more appropriate to compare Kirk's casualty rate to units on exploratory, rather than combat, missions. It must be remembered that the Enterprise under Kirk's command participated in multiple combat missions against warships with superior weaponry (such as in the episodes \"Balance of Terror\" and \"The Doomsday Machine\" to name a few). \\nAlso note that while this trope was true in a strictly numerical sense for the original Star Trek series (25 crew member died with red shirts on, 10 with gold, and 8 with blue), it is not true in terms of the percentage of red shirts shown. In percentages to total crew, 10 percent of red shirts died, against 18 percent of gold shirts.\\nIn mass quantities, they make up the Red Shirt Army. Frequently overlaps with Men Are the Expendable Gender and Black Dude Dies First.\\nCompare to The Worf Effect (a strong character is defeated to show the enemy's strength), Sacrificial Lion (a strong and important character is killed to show the enemy's strength or seriousness), The World's Expert on Getting Killed, Retirony, Mauve Shirt, Sacrificial Lamb, Disposable Sex Worker, Anyone Can Die, Little Dead Riding Hood, C-List Fodder, A Million Is a Statistic, and Monster Munch.\\nContrast Plot Armor and Red Herring Shirt.\\nFor John Scalzi's novel Redshirts, which deconstructs this trope, go to Literature.Redshirts, and for the sci-fi social networking simulator, go to VideoGame.Redshirt.\\nHas nothing to do with the guerrilla army of Italian nationalist leader Giuseppe Garibaldi.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>496</th>\n",
       "      <td>t30220</td>\n",
       "      <td>Bioaugmentation</td>\n",
       "      <td>The biological counterpart to Cyborging: instead of altering humans to add machine parts, the geeks in the lab-coats change the flesh, blood and DNA.\\nGenetic Engineering and stem cell research opens new avenues in the way science and Sci-Fi writers can create modified humans to suit any environment. You could think of them as artificially created Mutants, but \"artificially created\" is the key to the definition. And we mean created on purpose, not by freak accident like being bombarded with gamma radiation or other exposure to Green Rocks.\\nOn the mild end, this can entail things that, if current medical science keeps trending the way it does in Real Life, may not sound all that fantastic in a few decades: limb grafts, biological prosthetics and less visible gene mod enhancements that make people smarter, stronger and better at stuff. Further down the fantastic scale you start to get genetic superpowers, quick and easy drastic cosmetic changes from plastic surgery on up (want green skin and spots? No problem!), extra limbs and stuff that makes humans look a lot more like aliens than most Human Aliens.\\nCan be combined with Cyborging for even more body mod fun. Certain types of Functional Magic may accomplish the same thing but many or most of them are better listed under Mix And Match Creature or Hybrid Monster, and of those remaining, should be looked at with Clarke's Third Law in mind. Is a newer trope than cyborging because Genetic Engineering Is the New Nuke (and shares some but not all examples with it). Also see Organic Technology, LEGO Genetics, Magic Genetics, Superpowerful Genetics, Mix-and-Match Critters, Art Major Biology, Designer Babies, Bio Punk, and its Super-Trope, Transhuman. Can overlap with the Stock Superpower of Biomanipulation.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>497</th>\n",
       "      <td>t30286</td>\n",
       "      <td>XenoFiction</td>\n",
       "      <td>A few works of Science Fiction and Fantasy (and, in some cases even realistic fiction) take the Point of View of normal animals, Intellectual Animals, Intelligent Gerbils, or Starfish Aliens. More than that, the creators of such stories take great pains to think through what it would actually be like to be a rabbit, a dolphin, or a giant betentacled being who smells colour.\\nA good rule of thumb for figuring out if something is in this genre or not: if you can replace the non-humans with (maybe superpowered) humans without too much trouble, it's probably not Xenofiction: Beast Fables and works about Funny Animals are, in general, not examples. Likewise in many a Mouse World, size is often the only major difference between humans and non-humans; however, there is Xenofiction featuring non-humans with great differences in size.\\nIf it's taking place under the nose of humans, there may or may not have a Masquerade, and humans will probably either be bastards or eldritch abominations. If humans are taking place under the nose of it, you may have Humans Are Interesting. Xenofiction usually explores Bizarre Alien Psychology.\\nThis is also common in fanfics of Mon series where the mon is the POV character.\\nSee Xenofictional Literature for an index of Xenofiction.\\nContrast Most Writers Are Human. Compare to Narrative-Driven Nature Documentary for nature documentaries that anthropomorphise the animals.\\nNot to be confused with Xenafication, or the Xeno games.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>498</th>\n",
       "      <td>t30470</td>\n",
       "      <td>RecountedByTheMainCharacters</td>\n",
       "      <td>Sometimes, \"this is a true story\" is part of the fiction.\\nOnce in a while a really well written story can feel so real that you begin to wonder if it might not be based on a true story. Occasionally this is actually the case, or supposedly so, but there are times when an author (etc) will go right out of their way to create greater immersion in their work by claiming that their very obviously fictional and fantastic world is in some way real. Usually they claim that they didn't come up with the story, rather it was recounted to them by the actual main characters (or some other witness), often physically, but sometimes by phone or magic. Other times they will claim that they found the account in the form of a diary and novelised it, or, if it is a film, that it comprises found footage or a mixture of found footage and Dramatisation.\\nAnother common method is to claim that the book was written as a testimony (or confession) to actual events - possibly the most notable example of this is The Guild of Specialists trilogy, which takes the love that boys' annuals have for intricate diagrams and maps to its absolute extreme and fabricates not only a plethora of large diagrams, maps, and sketches, but photos and objects. In this version the author pretends they are simply publishing something that someone else has written - this often takes the form of a novelization of a diary or a set of notebooks. Other methods include accounts by secondary characters and so on. This trope, a staple of children's books and fantastic tales, often features an Author Avatar or even instances of From Beyond the Fourth Wall or other strangeness and may be said to be translated from accounts of what happened or books written by the characters and never actually communicated in person.\\nIn all these cases, however, it is considered canon that the author is repeating a story that is in fact true, if only to a certain degree. One of the people the story is about may even be the author themselves.\\nCompare ...And That Little Girl Was Me, and Based on a Great Big Lie.\\nAbsolutely not to be confused with Literary Agent Hypothesis, when fans think that maybe the story is actually real, or like to think it is, but don't have any support from canon or Word of God. Also not to be confused with A True Story in My Universe, for In-Universe examples. Often ties in with Author Avatar and may involve an admitted Unreliable Narrator.\\nNote: This trope only applies to canon and Word of God examples in fictional works. In-Universe examples go in A True Story in My Universe. Pure Fanon examples go in Literary Agent Hypothesis.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>499</th>\n",
       "      <td>t30487</td>\n",
       "      <td>DarkSkinnedRedHead</td>\n",
       "      <td>Like most character design decisions in anime, colors are often chosen to reflect personalities more than to be realistic depictions. Although dark skin is not a particularly alien choice, it's more often used to reflect a bouncy, energetic, fiery personality than actual nationality. This probably has roots back to the ideal of Japanese beauty including pale skin, as opposed to the chain of association with tanned people being outdoorsy, and thus lively and adventurous. Add in Western associations with red hair, and you get a boisterous adventure-seeker. Bonus points if they have gold eyes.\\nNaturally, this includes Hard Drinking Party Girls and Tricksters of both genders. The most common combination with this skin tone is a vibrant and noticeable light color. Red is very popular, associated with all sorts of playful Japanese spirits as well as the kabuki wig, so it's also a common (and deliberately unusual looking) palette choice for demons and Cat Girls. Red's a popular hair color for other reasons as well, of course.\\nNote, dark skinned characters with dark hair (olive, purple, black) tend to have muted personalities.\\nThis trope can be Truth in Television thanks to dye; notably, in Asian countries like Pakistan, dyeing hair red with henna dye is common. (Natural redheads in Pakistan — and there are many — usually have light skin.) Dark skin with natural red hair is often seen in Oceania, especially Vanuatu and Papua New Guinea. In other places, you generally only see natural red hair in cases of Mixed Ancestry. For example, the Afro-American activist Malcolm X earned the nickname \"Detroit Red\" due to the reddish tinge in his hair that he inherited from a Scottish ancestor. Likewise, basketball player Blake Griffin has tan skin and red hair due to having a black father and white redhead mother.\\nSister trope to Dark-Skinned Blond. Contrast Eerie Pale-Skinned Brunette and Raven Hair, Ivory Skin. Also quite likely to be Ambiguously Brown. But sometimes. You Gotta Have Blue Hair.</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>500 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    TropeID                         Trope  \\\n",
       "0    t00079                AbusiveParents   \n",
       "1    t00190                    ActionGirl   \n",
       "2    t00263        AdaptationDistillation   \n",
       "3    t00265           AdaptationExpansion   \n",
       "4    t00270          AdaptationNameChange   \n",
       "..      ...                           ...   \n",
       "495  t30000                     Redshirts   \n",
       "496  t30220               Bioaugmentation   \n",
       "497  t30286                   XenoFiction   \n",
       "498  t30470  RecountedByTheMainCharacters   \n",
       "499  t30487            DarkSkinnedRedHead   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                Description  \n",
       "0    Parents are supposed to be the protectors of children, but these parents are either so damaged themselves that they can't do the job, greedy or villainous to the point that they never had any interest in doing the job properly, or would rather use the child as a means to an end. Sometimes they're just sadistic assholes.\\nThis includes parents who are emotionally, verbally, physically, or mentally abusive, or who neglectfully allow their children to be abused by others if they don't abuse the child themselves; sexual abuse is typically treated as a special kind of evil. Sometimes, the abuse at the hands of their parents becomes a Freudian Excuse for a villain. Other times, the character manages to not grow up broken, bitter, and hateful, and instead a different and better person than the upbringing would incline one to think. Troubling Unchildlike Behavior is often a tell-tale sign that things are not right at home.\\nAbusive Parents are commonplace in fairy tales and Classical Mythology which makes this trope Older Than Feudalism. Note that The Brothers Grimm, when they collected European fairy tales, were uncomfortable with the idea of Abusive Parents and so frequently changed the Abusive Parents in the traditional stories into abusive step parents.\\nSometimes, a parent will go as far as to kill the child in question, in which case this is Offing the Offspring. In other ways cases, the parent's abuse occasionally drives the offspring to snap, commit Revenge and finally kill them, thus becoming a Self-Made Orphan. Although other times, the parents end up as a\\nKarma Houdini. Calling the Old Man Out occurs when a fed-up child retaliates with a \"The Reason You Suck\" Speech. If the child gets out of the broken family and forms healthy friendships, but reacts badly when their abusive parents show up again, well, Friends Are Chosen, Family Aren't.\\nBear in mind that not everyone agrees on the line between actual abuse and merely heavy-handed parenting (or even normal parenting). Is Moving the Goalposts merely inspiring the child to achieve more, or the most insidious form of abuse to instill mistrust and paranoia to the children? Some include spanking as abuse; others think it's appropriate given certain guidelines. Some believe it's okay to make a kid go without a meal (they won't starve that easily); others disagree. Making a kid miss a friend's birthday sleepover — is that emotional abuse? Raising a kid without exposure to TV? Telling your daughter she's getting fat? A little friendly name-calling? There's a line here somewhere, but not everyone agrees on where it is.\\nIf a parent has just dumped the child, for whatever reason, that's Parental Abandonment; if they aren't paying attention, that's Parental Neglect. If the parents refuse to discipline their kids, they are Pushover Parents. Contrast Mama Bear or Papa Wolf (where others abuse the children and the parents abuse the abusers), and the more extreme variant of Knight Templar Parent, where the abusive parent is violently overprotective. Abusive Precursors can be considered this on a metaphorical level. See Hilariously Abusive Childhood for when this is cranked up to absurd levels and played for laughs. Black Comedy is often connected in the comedic aspect of it, and a Big, Screwed-Up Family may be involved if it is adult comedy. In keeping with the above note, some may call the show on it and say Dude, Not Funny!. See Evil Matriarch and Archnemesis Dad for characters who are beyond abusive and outright evil. For parents who are mostly abused by their children, see Pushover Parents. Abusive-type parents are mostly wanted criminals being chased by the law.\\nWhile they do not have to be the child's actual, technical parents to be part of this trope, it's pretty important that they are closely related and live together, like a Wicked Stepmother or an Evil Uncle taking care of the Parentally Deprived. After all, it's much more disgusting that somebody related to the child could bring themselves to hurt them, rather than a mere foster family.\\nThe polar opposite, of course, are Good Parents.\\nNOTE: Please do not use this trope for complaining about parents you don't like.\\nUnfortunately, Abusive Parents are Truth in Television and a very sensitive topic for many, so please no Real Life examples - it is sufficient to say they do exist and they are all too common.  \n",
       "1                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          An Action Girl is a female badass who is tough and kicks butt. Damsel in Distress? Not her. She's featured in far more than the Designated Girl Fight. She faces dangerous foes and deadly obstacles, and she wins.\\nFor the longest time in many cultures, Double Standards in both fiction and real life meant that when it came to action and fighting, guys definitely outnumbered girls. Men Act, Women Are was the rule of thumb, which led to the Action Girl being a subversion of what was acceptable. As society has marched on, this view has faded in some media, leading to stories where Action Girls become the norm rather than the exception, such as stories set in a World of Badass, and especially a World of Action Girls.\\nThe broad Action Girl concept can take many forms. Faux Action Girl is a case of presenting a character as this, only for them not to live up to the standards of this trope. Dark Action Girl is the villainous variety, and Affirmative Action Girl is a cast addition intended to balance out gender ratios that typically also falls under this trope. Less action-y versions include You Go, Girl! and Plucky Girl.\\nThis is somewhat of a Cyclic Trope, with Action Girls often having surges in popularity in the 1940s, 1970s, 1990s and 2010s, probably not coincidentally times when the women's movement increased in prominence.\\nSee below for even more subtropes and other variants.\\nNot to be confused with Action Girl, a comic series about an Action Girl. For advice on how to write these types of characters, see Write An Action Girl.\\n\\nGenres and worlds populated by Action Girls:  \n",
       "2                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  Some adaptations take a complex character or situation and greatly simplify it, removing elements the producer believed to be unnecessary. This effect is more common when adapting from a long-running series, especially if it hasn't had a singular vision over the years.\\nCompare Compressed Adaptation, which deals more along the lines of a direct adaptation and is about all the stuff that is cut out along the way to make this trope happen.\\nContrast with Pragmatic Adaptation: in a distillation, a complex story is simplified, without much substantive change. In a Pragmatic Adaptation, the story is changed with the shift in medium. Also, contrast Adaptation Expansion, which tacks on more content but retains the original source material. When a story element is removed but its effects aren't, that's Adaptation Explanation Extrication.\\nThis isn't about adaptations that are better than the original. This is about elements that are simplified or removed in order to make an adaptation shorter.  \n",
       "3                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            This is the complete opposite of Compressed Adaptation. It occurs when a short, very simple tale is given a much bigger adaptation, typically as a feature-length film. To be brought up to feature length, the storyline will have to be padded with some new stuff — a lot of new stuff. Cue Alternative Character Interpretations that require elaborate backstories, minor characters given much larger parts, completely new characters, and sometimes Plot Holes, Plot Tumors, and a triple dozen subplots that were not in the original work.\\nThis has a tendency to make the story unrecognizable as a retelling of the original. In some instances, the original story will end up as one small part of a much larger, more convoluted story. This will usually be the climax, in which case the film essentially gave you an hour or more of Back Story. This most often happens with movies based on novellas, short stories, video games or children's books. In the children's books instance, this can lead to the introduction of Darker and Edgier into a normally benign story or the reintroduction of elements lost to Disneyfication of classic stories.\\nWhile this trope is often associated with the fan complaint \"They Changed It, Now It Sucks!!\", it can be done well. Remember that Tropes Are Tools and extensions for adaptations might need to happen to fill the required runtime, just going about it right is the problem most productions face.\\nCompare Expanded Universe. See also Adaptation Distillation, Compressed Adaptation, Humble Beginnings, Patchwork Story, Not His Sled, Updated Re Release. An adaptation In Name Only goes even further than this, throwing out the original plot and making things up out of whole cloth.  \n",
       "4                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                Alice Andrews in the books becomes Alice Allen in The Movie then Annie Adams in The Series.\\nThis is when a character or group has a name change between adaptations.\\nThe Name Change of The Character can be their first name, last name or both that also includes super alter egos.\\nThe key to this trope is that between adaptations, the character must clearly be the same. If they are a Suspiciously Similar Substitute or a Composite Character, then this trope does not apply.\\nThis can sometimes be an Enforced Trope. Production studios retain a team of lawyers whose job it is to ensure that every name in every work is shared either by many people or by no people in the area that the work is being produced, a process known as \"clearance\". This is to avoid a lawsuit: if there's exactly one person named Darkness von Gothick, you're just begging for him to sue you for portraying him as a villain. But names that are common in one place and time might be more rare in another, resulting in a name being cleared for use when The Movie was filmed in Hollywood fifty years ago but not when The Series is filmed in Vancouver today.\\nThis is also very common in foreign adaptations, where names are changed to either reflect the local culture, or to give the show its own identity. It may also accompany a Race Lift, Adaptational Nationality change, or Gender Flip, with a character's name reflecting his or her new ethnicity or gender (usually).\\nThis trope only applies when the adaptation is in the same language as the original. When a character's name changes in a translation, that's Dub Name Change. If a character didn't originally have any name then they're Named by the Adaptation. If it's Based on a True Story, see Roman à Clef.  \n",
       "..                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      ...  \n",
       "495                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                This is the Good Counterpart of Evil Minions and Mooks — set filler for our heroes' side. Their purpose is almost exclusively to give the writers someone to kill who isn't a main character, although they can also serve as Spear Carriers. In a series where The Main Characters Do Everything, if you suddenly see someone else who you've never seen before involved in the main story, they are probably Redshirts.\\nThey are used to show how the monster works, and demonstrate that it is indeed a deadly menace, without having to lose anyone important. Expect someone to say \"He's Dead, Jim\", lament this \"valued crew member's senseless death\", and then promptly forget him. Security personnel in general fall victim to the worst shade of this trope, as most of the time their deaths aren't even acknowledged at all; according to Hollywood, you could walk into a bank and shoot a security guard right in the face without anyone making a fuss. If you shot anyone else afterward, the headline would just read \"Bank Customers Killed\".\\nPlease note: this Trope is actually very inaccurate when you compare it to Real Life. If you were to watch every episode of Star Trek: The Original Series, count the number of casualties that the Enterprise had, and then compare that to an actual military, you'd see that Kirk's record as a leader in this regard is excellent, far better than any general in U.S. history. Even war heroes like George Washington and Dwight D. Eisenhower had proportionately more casualties among their troops.note It has been suggested that it would be more appropriate to compare Kirk's casualty rate to units on exploratory, rather than combat, missions. It must be remembered that the Enterprise under Kirk's command participated in multiple combat missions against warships with superior weaponry (such as in the episodes \"Balance of Terror\" and \"The Doomsday Machine\" to name a few). \\nAlso note that while this trope was true in a strictly numerical sense for the original Star Trek series (25 crew member died with red shirts on, 10 with gold, and 8 with blue), it is not true in terms of the percentage of red shirts shown. In percentages to total crew, 10 percent of red shirts died, against 18 percent of gold shirts.\\nIn mass quantities, they make up the Red Shirt Army. Frequently overlaps with Men Are the Expendable Gender and Black Dude Dies First.\\nCompare to The Worf Effect (a strong character is defeated to show the enemy's strength), Sacrificial Lion (a strong and important character is killed to show the enemy's strength or seriousness), The World's Expert on Getting Killed, Retirony, Mauve Shirt, Sacrificial Lamb, Disposable Sex Worker, Anyone Can Die, Little Dead Riding Hood, C-List Fodder, A Million Is a Statistic, and Monster Munch.\\nContrast Plot Armor and Red Herring Shirt.\\nFor John Scalzi's novel Redshirts, which deconstructs this trope, go to Literature.Redshirts, and for the sci-fi social networking simulator, go to VideoGame.Redshirt.\\nHas nothing to do with the guerrilla army of Italian nationalist leader Giuseppe Garibaldi.  \n",
       "496                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     The biological counterpart to Cyborging: instead of altering humans to add machine parts, the geeks in the lab-coats change the flesh, blood and DNA.\\nGenetic Engineering and stem cell research opens new avenues in the way science and Sci-Fi writers can create modified humans to suit any environment. You could think of them as artificially created Mutants, but \"artificially created\" is the key to the definition. And we mean created on purpose, not by freak accident like being bombarded with gamma radiation or other exposure to Green Rocks.\\nOn the mild end, this can entail things that, if current medical science keeps trending the way it does in Real Life, may not sound all that fantastic in a few decades: limb grafts, biological prosthetics and less visible gene mod enhancements that make people smarter, stronger and better at stuff. Further down the fantastic scale you start to get genetic superpowers, quick and easy drastic cosmetic changes from plastic surgery on up (want green skin and spots? No problem!), extra limbs and stuff that makes humans look a lot more like aliens than most Human Aliens.\\nCan be combined with Cyborging for even more body mod fun. Certain types of Functional Magic may accomplish the same thing but many or most of them are better listed under Mix And Match Creature or Hybrid Monster, and of those remaining, should be looked at with Clarke's Third Law in mind. Is a newer trope than cyborging because Genetic Engineering Is the New Nuke (and shares some but not all examples with it). Also see Organic Technology, LEGO Genetics, Magic Genetics, Superpowerful Genetics, Mix-and-Match Critters, Art Major Biology, Designer Babies, Bio Punk, and its Super-Trope, Transhuman. Can overlap with the Stock Superpower of Biomanipulation.  \n",
       "497                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    A few works of Science Fiction and Fantasy (and, in some cases even realistic fiction) take the Point of View of normal animals, Intellectual Animals, Intelligent Gerbils, or Starfish Aliens. More than that, the creators of such stories take great pains to think through what it would actually be like to be a rabbit, a dolphin, or a giant betentacled being who smells colour.\\nA good rule of thumb for figuring out if something is in this genre or not: if you can replace the non-humans with (maybe superpowered) humans without too much trouble, it's probably not Xenofiction: Beast Fables and works about Funny Animals are, in general, not examples. Likewise in many a Mouse World, size is often the only major difference between humans and non-humans; however, there is Xenofiction featuring non-humans with great differences in size.\\nIf it's taking place under the nose of humans, there may or may not have a Masquerade, and humans will probably either be bastards or eldritch abominations. If humans are taking place under the nose of it, you may have Humans Are Interesting. Xenofiction usually explores Bizarre Alien Psychology.\\nThis is also common in fanfics of Mon series where the mon is the POV character.\\nSee Xenofictional Literature for an index of Xenofiction.\\nContrast Most Writers Are Human. Compare to Narrative-Driven Nature Documentary for nature documentaries that anthropomorphise the animals.\\nNot to be confused with Xenafication, or the Xeno games.  \n",
       "498                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    Sometimes, \"this is a true story\" is part of the fiction.\\nOnce in a while a really well written story can feel so real that you begin to wonder if it might not be based on a true story. Occasionally this is actually the case, or supposedly so, but there are times when an author (etc) will go right out of their way to create greater immersion in their work by claiming that their very obviously fictional and fantastic world is in some way real. Usually they claim that they didn't come up with the story, rather it was recounted to them by the actual main characters (or some other witness), often physically, but sometimes by phone or magic. Other times they will claim that they found the account in the form of a diary and novelised it, or, if it is a film, that it comprises found footage or a mixture of found footage and Dramatisation.\\nAnother common method is to claim that the book was written as a testimony (or confession) to actual events - possibly the most notable example of this is The Guild of Specialists trilogy, which takes the love that boys' annuals have for intricate diagrams and maps to its absolute extreme and fabricates not only a plethora of large diagrams, maps, and sketches, but photos and objects. In this version the author pretends they are simply publishing something that someone else has written - this often takes the form of a novelization of a diary or a set of notebooks. Other methods include accounts by secondary characters and so on. This trope, a staple of children's books and fantastic tales, often features an Author Avatar or even instances of From Beyond the Fourth Wall or other strangeness and may be said to be translated from accounts of what happened or books written by the characters and never actually communicated in person.\\nIn all these cases, however, it is considered canon that the author is repeating a story that is in fact true, if only to a certain degree. One of the people the story is about may even be the author themselves.\\nCompare ...And That Little Girl Was Me, and Based on a Great Big Lie.\\nAbsolutely not to be confused with Literary Agent Hypothesis, when fans think that maybe the story is actually real, or like to think it is, but don't have any support from canon or Word of God. Also not to be confused with A True Story in My Universe, for In-Universe examples. Often ties in with Author Avatar and may involve an admitted Unreliable Narrator.\\nNote: This trope only applies to canon and Word of God examples in fictional works. In-Universe examples go in A True Story in My Universe. Pure Fanon examples go in Literary Agent Hypothesis.  \n",
       "499                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          Like most character design decisions in anime, colors are often chosen to reflect personalities more than to be realistic depictions. Although dark skin is not a particularly alien choice, it's more often used to reflect a bouncy, energetic, fiery personality than actual nationality. This probably has roots back to the ideal of Japanese beauty including pale skin, as opposed to the chain of association with tanned people being outdoorsy, and thus lively and adventurous. Add in Western associations with red hair, and you get a boisterous adventure-seeker. Bonus points if they have gold eyes.\\nNaturally, this includes Hard Drinking Party Girls and Tricksters of both genders. The most common combination with this skin tone is a vibrant and noticeable light color. Red is very popular, associated with all sorts of playful Japanese spirits as well as the kabuki wig, so it's also a common (and deliberately unusual looking) palette choice for demons and Cat Girls. Red's a popular hair color for other reasons as well, of course.\\nNote, dark skinned characters with dark hair (olive, purple, black) tend to have muted personalities.\\nThis trope can be Truth in Television thanks to dye; notably, in Asian countries like Pakistan, dyeing hair red with henna dye is common. (Natural redheads in Pakistan — and there are many — usually have light skin.) Dark skin with natural red hair is often seen in Oceania, especially Vanuatu and Papua New Guinea. In other places, you generally only see natural red hair in cases of Mixed Ancestry. For example, the Afro-American activist Malcolm X earned the nickname \"Detroit Red\" due to the reddish tinge in his hair that he inherited from a Scottish ancestor. Likewise, basketball player Blake Griffin has tan skin and red hair due to having a black father and white redhead mother.\\nSister trope to Dark-Skinned Blond. Contrast Eerie Pale-Skinned Brunette and Raven Hair, Ivory Skin. Also quite likely to be Ambiguously Brown. But sometimes. You Gotta Have Blue Hair.  \n",
       "\n",
       "[500 rows x 3 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tropes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "334aa167-3458-42f1-9cf0-d368bcce9792",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Title, Trope, Example, trope_id, title_id <- important fields\n",
    "def get_film_examples(trope_list):\n",
    "    df1 = pd.read_csv('../dataset/TVTropesData/film_tropes.csv', index_col=0, dtype={'trope_id': str, \"title_id\": str, \"Example\": str, \"Trope\": str, \"Title\": str})\n",
    "    # df2 = pd.read_csv('../dataset/TVTropesData/film_imdb_match.csv', index_col=0, dtype={'trope_id': str, \"title_id\": str, \"Example\": str, \"Trope\": str, \"Title\": str})\n",
    "    part1 = df1[df1['trope_id'].isin(trope_list)]#['Title', 'Trope', 'Example', 'trope_id', 'title_id']\n",
    "    # part2 = df2[df2['trope_id'].isin(trope_list)]#['Title', 'Trope', 'Example', 'trope_id', 'title_id']\n",
    "    print(part1.head())\n",
    "    # print(part2.head())\n",
    "    part1.Example = part1.Example.str.strip(' \\t\\n\\r')\n",
    "    # part2.Example = part2.Example.str.strip(' \\t\\n\\r')\n",
    "    #concatanate and return\n",
    "    # film_examples = pd.concat([part1[['Example', 'trope_id']], part2[['Example', 'trope_id']]], ignore_index=True)\n",
    "    film_examples = part1[['Example', 'trope_id']].reset_index()\n",
    "    return film_examples\n",
    "\n",
    "def get_lit_examples(trope_list):\n",
    "    df1 = pd.read_csv('../dataset/TVTropesData/lit_tropes.csv', index_col=0, dtype={'trope_id': str, \"title_id\": str, \"Example\": str, \"Trope\": str, \"Title\": str})#.dropna()\n",
    "    # df2 = pd.read_csv('../dataset/TVTropesData/lit_goodreads_match.csv', index_col=0, dtype={'trope_id': str, \"title_id\": str, \"Example\": str, \"Trope\": str, \"Title\": str})#.dropna()\n",
    "    part1 = df1[df1['trope_id'].isin(trope_list)]#['Title', 'Trope', 'Example', 'trope_id', 'title_id']\n",
    "    # part2 = df2[df2['trope_id'].isin(trope_list)]#['Title', 'Trope', 'Example', 'trope_id', 'title_id']\n",
    "    print(part1.head())\n",
    "    # print(part2.head())\n",
    "    part1.Example = part1.Example.str.strip(' \\t\\n\\r')\n",
    "    # part2.Example = part2.Example.str.strip(' \\t\\n\\r')\n",
    "    # lit_examples = pd.concat([part1[['Example', 'trope_id']], part2[['Example', 'trope_id']]], ignore_index=True)\n",
    "    lit_examples = part1[['Example', 'trope_id']].reset_index()\n",
    "    return lit_examples\n",
    "\n",
    "# i do not think these would be that relevant. but thats just a conjecture, needs to be checked\n",
    "def get_tv_examples():\n",
    "    pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ba39c3c6-8ab6-440c-ad68-3ffd55627d53",
   "metadata": {},
   "outputs": [],
   "source": [
    "tropes_list = tropes[\"TropeID\"].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "0dc3ddc3-d6cb-4beb-a1b6-b999b80ff14a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               Title                     Trope  \\\n",
      "10  ABadDayForVoodoo              TheChosenOne   \n",
      "13  ABadDayForVoodoo  ArsonMurderAndJaywalking   \n",
      "15  ABadDayForVoodoo              TheChosenOne   \n",
      "16  ABadDayForVoodoo                  AntiHero   \n",
      "17  ABadDayForVoodoo                  WhamLine   \n",
      "\n",
      "                                                                                                                                                                                                                                                                           Example  \\\n",
      "10   In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.\\n   \n",
      "13                                                              When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".   \n",
      "15                         Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
      "16                                                                                                                                              Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.   \n",
      "17                                                                                      Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .   \n",
      "\n",
      "   trope_id title_id  \n",
      "10   t22727     lit1  \n",
      "13   t01294     lit1  \n",
      "15   t22727     lit1  \n",
      "16   t01080     lit1  \n",
      "17   t25818     lit1  \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\joe\\AppData\\Local\\Temp\\ipykernel_14012\\2331974135.py:23: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  part1.Example = part1.Example.str.strip(' \\t\\n\\r')\n"
     ]
    }
   ],
   "source": [
    "lit_examples = get_lit_examples(tropes_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "f2774903-3dba-41b6-aae1-b5bb883f0af2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "           Title                    Trope  \\\n",
      "3   ABBATheMovie  GettingCrapPastTheRadar   \n",
      "7   ABBATheMovie           DeadpanSnarker   \n",
      "21  ABBATheMovie     ContrivedCoincidence   \n",
      "24  ABBATheMovie             MythologyGag   \n",
      "27  ABBATheMovie               ButtMonkey   \n",
      "\n",
      "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 Example  \\\n",
      "3                                                                                                                                                                                                                                                                                                                                    During Ashley's fantasy sequence in which he imagines Agnetha and Frida being in love with him, at one point Agnetha and Frida are both draped over him in an  Almost Kiss  before it cuts away to another scene, which  really  looks like it's  cutting away  from a  threesome .   \n",
      "7                                                                                       The band have their moments.  One of the crowning examples is Agnetha's reaction after discovering that one news headline about the band reads \"Agnetha's Bottom Tops Show\"  Agnetha   Oh my God! Don't they have bottoms in Australia?!  Then there's her reaction immediately after, as Björn reads the article in question  Björn    [reading from the article]  \"A middle-aged man beside me said, with feeling, 'It's  really  something worth seeing!'\"  Agnetha   Hey, at least I did  something  for the show, didn't I!   \n",
      "21                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         Ashley finally obtains his interview when he steps into an elevator in which ABBA just also happens to be in.   \n",
      "24    The country song heard playing at the start of the film when Ashley is in the radio studio, \"Please Change Your Mind\", is a song composed by Björn & Benny in the early 1970s back when they were a duo, and is performed by Nashville Train, a Swedish country band composed of members of ABBA's own backing band.  At one point during Ashley's fantasy sequence, he imagines himself inside an Old West saloon with the band. This seems to be a reference to one of the band's  early TV appearances  as Festfolket, when the band were more of a comedy troupe, which had them appear on an Old West saloon.   \n",
      "27                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               Ashley seems unable to catch a break during his quest for an interview.   \n",
      "\n",
      "   trope_id title_id  \n",
      "3    t09016       f0  \n",
      "7    t05235       f0  \n",
      "21   t04387       f0  \n",
      "24   t14940       f0  \n",
      "27   t03051       f0  \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\joe\\AppData\\Local\\Temp\\ipykernel_14012\\2331974135.py:9: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  part1.Example = part1.Example.str.strip(' \\t\\n\\r')\n"
     ]
    }
   ],
   "source": [
    "film_examples = get_film_examples(tropes_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "454ff7ec-2a21-423b-af1b-0f5407c78da4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Example</th>\n",
       "      <th>trope_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13</td>\n",
       "      <td>When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".</td>\n",
       "      <td>t01294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15</td>\n",
       "      <td>Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16</td>\n",
       "      <td>Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.</td>\n",
       "      <td>t01080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17</td>\n",
       "      <td>Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .</td>\n",
       "      <td>t25818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101518</th>\n",
       "      <td>685032</td>\n",
       "      <td>Eugene Palette's \"Fray Felipe\" in the Tyrone Power version is suspiciously similar to his Friar Tuck in  The Adventures of Robin Hood .</td>\n",
       "      <td>t07317</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101519</th>\n",
       "      <td>685038</td>\n",
       "      <td>The narrator of Zuleika Dobson appears to be a hold-over from the Victorian use of this trope (i.e. an opinionated narrator whose in-universe/out-of-universe status is nebulous). Then he's revealed halfway through the book to be a historian whom the Greek muse of history has granted the ability to be invisibly, intangibly present for the events of the story so that he can write a history book that's more like a novel (and therefore less boring).</td>\n",
       "      <td>t12755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101520</th>\n",
       "      <td>685042</td>\n",
       "      <td>Reflecting on the fatal events of the novel, Zuleika's chief regret is that it won't be possible to prove that each and every death was for love of her.  She decides that she'll have to do better next time  Cambridge, here she comes.</td>\n",
       "      <td>t11920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101521</th>\n",
       "      <td>685045</td>\n",
       "      <td>By the time the Duke thinks better of his plan of suicide, it's too late  the traditional omens of a death in his family have already taken place.</td>\n",
       "      <td>t26457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101522</th>\n",
       "      <td>685049</td>\n",
       "      <td>Zuleika has never loved, because she cannot love anyone who adores her. And  everyone  adores her.</td>\n",
       "      <td>t20764</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>101523 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         index  \\\n",
       "0           10   \n",
       "1           13   \n",
       "2           15   \n",
       "3           16   \n",
       "4           17   \n",
       "...        ...   \n",
       "101518  685032   \n",
       "101519  685038   \n",
       "101520  685042   \n",
       "101521  685045   \n",
       "101522  685049   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                                                                                                                                                                                  Example  \\\n",
       "0                                                                                                                                                                                             In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "1                                                                                                                                                                                                                                                      When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".   \n",
       "2                                                                                                                                                                                                                 Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "3                                                                                                                                                                                                                                                                                                                                      Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.   \n",
       "4                                                                                                                                                                                                                                                                              Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .   \n",
       "...                                                                                                                                                                                                                                                                                                                                                                                                                                                                   ...   \n",
       "101518                                                                                                                                                                                                                                                                                                                            Eugene Palette's \"Fray Felipe\" in the Tyrone Power version is suspiciously similar to his Friar Tuck in  The Adventures of Robin Hood .   \n",
       "101519  The narrator of Zuleika Dobson appears to be a hold-over from the Victorian use of this trope (i.e. an opinionated narrator whose in-universe/out-of-universe status is nebulous). Then he's revealed halfway through the book to be a historian whom the Greek muse of history has granted the ability to be invisibly, intangibly present for the events of the story so that he can write a history book that's more like a novel (and therefore less boring).   \n",
       "101520                                                                                                                                                                                                                          Reflecting on the fatal events of the novel, Zuleika's chief regret is that it won't be possible to prove that each and every death was for love of her.  She decides that she'll have to do better next time  Cambridge, here she comes.   \n",
       "101521                                                                                                                                                                                                                                                                                                                 By the time the Duke thinks better of his plan of suicide, it's too late  the traditional omens of a death in his family have already taken place.   \n",
       "101522                                                                                                                                                                                                                                                                                                                                                                 Zuleika has never loved, because she cannot love anyone who adores her. And  everyone  adores her.   \n",
       "\n",
       "       trope_id  \n",
       "0        t22727  \n",
       "1        t01294  \n",
       "2        t22727  \n",
       "3        t01080  \n",
       "4        t25818  \n",
       "...         ...  \n",
       "101518   t07317  \n",
       "101519   t12755  \n",
       "101520   t11920  \n",
       "101521   t26457  \n",
       "101522   t20764  \n",
       "\n",
       "[101523 rows x 3 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lit_examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "dfc6c7ba-435b-4bef-b9db-9a709631ad90",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "t20214    1431\n",
       "t13853    1103\n",
       "t29575    1058\n",
       "t02187     953\n",
       "t02336     788\n",
       "          ... \n",
       "t28705      27\n",
       "t00499      27\n",
       "t23221      27\n",
       "t30000      27\n",
       "t21795      26\n",
       "Name: trope_id, Length: 500, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lit_examples.trope_id.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "aa2cc7d0-34f6-46b8-913f-2554cf2968a6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Example</th>\n",
       "      <th>trope_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>During Ashley's fantasy sequence in which he imagines Agnetha and Frida being in love with him, at one point Agnetha and Frida are both draped over him in an  Almost Kiss  before it cuts away to another scene, which  really  looks like it's  cutting away  from a  threesome .</td>\n",
       "      <td>t09016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7</td>\n",
       "      <td>The band have their moments.  One of the crowning examples is Agnetha's reaction after discovering that one news headline about the band reads \"Agnetha's Bottom Tops Show\"  Agnetha   Oh my God! Don't they have bottoms in Australia?!  Then there's her reaction immediately after, as Björn reads the article in question  Björn    [reading from the article]  \"A middle-aged man beside me said, with feeling, 'It's  really  something worth seeing!'\"  Agnetha   Hey, at least I did  something  for the show, didn't I!</td>\n",
       "      <td>t05235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21</td>\n",
       "      <td>Ashley finally obtains his interview when he steps into an elevator in which ABBA just also happens to be in.</td>\n",
       "      <td>t04387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>24</td>\n",
       "      <td>The country song heard playing at the start of the film when Ashley is in the radio studio, \"Please Change Your Mind\", is a song composed by Björn &amp; Benny in the early 1970s back when they were a duo, and is performed by Nashville Train, a Swedish country band composed of members of ABBA's own backing band.  At one point during Ashley's fantasy sequence, he imagines himself inside an Old West saloon with the band. This seems to be a reference to one of the band's  early TV appearances  as Festfolket, when the band were more of a comedy troupe, which had them appear on an Old West saloon.</td>\n",
       "      <td>t14940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>27</td>\n",
       "      <td>Ashley seems unable to catch a break during his quest for an interview.</td>\n",
       "      <td>t03051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94356</th>\n",
       "      <td>758972</td>\n",
       "      <td>And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!</td>\n",
       "      <td>t12565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94357</th>\n",
       "      <td>758985</td>\n",
       "      <td>There's a Star Wars Expanded Universe comic featuring Luke's childhood friend Janek \"Tank\" Sunber, who'd joined the Empire, become a lieutenant, and ended up stationed on a planet of tribal aliens. The plot of that handful of comics is essentially Zulu, with Imperials desperately fighting wave after wave of aliens and being worn down.</td>\n",
       "      <td>t26776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94358</th>\n",
       "      <td>759016</td>\n",
       "      <td>Private Thomas, one of the Welsh soldiers in Zulu, who gets chewed out for taking care of a calf instead of being on duty early in the film.</td>\n",
       "      <td>t08978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94359</th>\n",
       "      <td>759046</td>\n",
       "      <td>According to Urban Legend, the Zulu warrior in Zulu who falls dying before his chief, and who makes a heartfelt declaration of loyalty as he does so, is actually saying something along the lines of \"I have an enormous penis\" in Swahilinotenote that Zulu people generally speak isiZulu, not Swahili.</td>\n",
       "      <td>t02282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94360</th>\n",
       "      <td>759063</td>\n",
       "      <td>Zwartboek (Black Book): Deals with the Dutch resistance in the Nazi-occupied Netherlands during World War II.</td>\n",
       "      <td>t12563</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>94361 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        index  \\\n",
       "0           3   \n",
       "1           7   \n",
       "2          21   \n",
       "3          24   \n",
       "4          27   \n",
       "...       ...   \n",
       "94356  758972   \n",
       "94357  758985   \n",
       "94358  759016   \n",
       "94359  759046   \n",
       "94360  759063   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  Example  \\\n",
       "0                                                                                                                                                                                                                                                                                                                                     During Ashley's fantasy sequence in which he imagines Agnetha and Frida being in love with him, at one point Agnetha and Frida are both draped over him in an  Almost Kiss  before it cuts away to another scene, which  really  looks like it's  cutting away  from a  threesome .   \n",
       "1                                                                                        The band have their moments.  One of the crowning examples is Agnetha's reaction after discovering that one news headline about the band reads \"Agnetha's Bottom Tops Show\"  Agnetha   Oh my God! Don't they have bottoms in Australia?!  Then there's her reaction immediately after, as Björn reads the article in question  Björn    [reading from the article]  \"A middle-aged man beside me said, with feeling, 'It's  really  something worth seeing!'\"  Agnetha   Hey, at least I did  something  for the show, didn't I!   \n",
       "2                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           Ashley finally obtains his interview when he steps into an elevator in which ABBA just also happens to be in.   \n",
       "3      The country song heard playing at the start of the film when Ashley is in the radio studio, \"Please Change Your Mind\", is a song composed by Björn & Benny in the early 1970s back when they were a duo, and is performed by Nashville Train, a Swedish country band composed of members of ABBA's own backing band.  At one point during Ashley's fantasy sequence, he imagines himself inside an Old West saloon with the band. This seems to be a reference to one of the band's  early TV appearances  as Festfolket, when the band were more of a comedy troupe, which had them appear on an Old West saloon.   \n",
       "4                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 Ashley seems unable to catch a break during his quest for an interview.   \n",
       "...                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   ...   \n",
       "94356                                                                                                                                                                                                                                                                                                                                                                                       And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!   \n",
       "94357                                                                                                                                                                                                                                                                    There's a Star Wars Expanded Universe comic featuring Luke's childhood friend Janek \"Tank\" Sunber, who'd joined the Empire, become a lieutenant, and ended up stationed on a planet of tribal aliens. The plot of that handful of comics is essentially Zulu, with Imperials desperately fighting wave after wave of aliens and being worn down.   \n",
       "94358                                                                                                                                                                                                                                                                                                                                                                                                                                                                        Private Thomas, one of the Welsh soldiers in Zulu, who gets chewed out for taking care of a calf instead of being on duty early in the film.   \n",
       "94359                                                                                                                                                                                                                                                                                                          According to Urban Legend, the Zulu warrior in Zulu who falls dying before his chief, and who makes a heartfelt declaration of loyalty as he does so, is actually saying something along the lines of \"I have an enormous penis\" in Swahilinotenote that Zulu people generally speak isiZulu, not Swahili.   \n",
       "94360                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       Zwartboek (Black Book): Deals with the Dutch resistance in the Nazi-occupied Netherlands during World War II.   \n",
       "\n",
       "      trope_id  \n",
       "0       t09016  \n",
       "1       t05235  \n",
       "2       t04387  \n",
       "3       t14940  \n",
       "4       t03051  \n",
       "...        ...  \n",
       "94356   t12565  \n",
       "94357   t26776  \n",
       "94358   t08978  \n",
       "94359   t02282  \n",
       "94360   t12563  \n",
       "\n",
       "[94361 rows x 3 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "film_examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a3b1dd78-9ec5-45b0-853e-57f7a9adfdc6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "t20214    2091\n",
       "t03644    1343\n",
       "t15984    1340\n",
       "t08393    1232\n",
       "t02187    1229\n",
       "          ... \n",
       "t24860       4\n",
       "t30220       3\n",
       "t26889       3\n",
       "t15494       1\n",
       "t28962       1\n",
       "Name: trope_id, Length: 496, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "film_examples.trope_id.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "3abade9f-2d14-454f-9722-1d27bcedac87",
   "metadata": {},
   "outputs": [],
   "source": [
    "lit_examples['example_type'] = \"lit\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "d4a30a22-02e1-4135-ae40-79c4ca8fe506",
   "metadata": {},
   "outputs": [],
   "source": [
    "film_examples['example_type'] = \"film\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "28a1d6df-9bde-436a-b6e4-703dc86c0727",
   "metadata": {},
   "outputs": [],
   "source": [
    "trope_examples = pd.concat([lit_examples, film_examples], ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "1854cf90-fcb8-48f8-8f3e-9ff72262e0ad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Example</th>\n",
       "      <th>trope_id</th>\n",
       "      <th>example_type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13</td>\n",
       "      <td>When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".</td>\n",
       "      <td>t01294</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15</td>\n",
       "      <td>Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16</td>\n",
       "      <td>Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.</td>\n",
       "      <td>t01080</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17</td>\n",
       "      <td>Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .</td>\n",
       "      <td>t25818</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195879</th>\n",
       "      <td>758972</td>\n",
       "      <td>And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!</td>\n",
       "      <td>t12565</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195880</th>\n",
       "      <td>758985</td>\n",
       "      <td>There's a Star Wars Expanded Universe comic featuring Luke's childhood friend Janek \"Tank\" Sunber, who'd joined the Empire, become a lieutenant, and ended up stationed on a planet of tribal aliens. The plot of that handful of comics is essentially Zulu, with Imperials desperately fighting wave after wave of aliens and being worn down.</td>\n",
       "      <td>t26776</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195881</th>\n",
       "      <td>759016</td>\n",
       "      <td>Private Thomas, one of the Welsh soldiers in Zulu, who gets chewed out for taking care of a calf instead of being on duty early in the film.</td>\n",
       "      <td>t08978</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195882</th>\n",
       "      <td>759046</td>\n",
       "      <td>According to Urban Legend, the Zulu warrior in Zulu who falls dying before his chief, and who makes a heartfelt declaration of loyalty as he does so, is actually saying something along the lines of \"I have an enormous penis\" in Swahilinotenote that Zulu people generally speak isiZulu, not Swahili.</td>\n",
       "      <td>t02282</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195883</th>\n",
       "      <td>759063</td>\n",
       "      <td>Zwartboek (Black Book): Deals with the Dutch resistance in the Nazi-occupied Netherlands during World War II.</td>\n",
       "      <td>t12563</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>195884 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         index  \\\n",
       "0           10   \n",
       "1           13   \n",
       "2           15   \n",
       "3           16   \n",
       "4           17   \n",
       "...        ...   \n",
       "195879  758972   \n",
       "195880  758985   \n",
       "195881  759016   \n",
       "195882  759046   \n",
       "195883  759063   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                                                                 Example  \\\n",
       "0                                                                            In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "1                                                                                                                                     When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".   \n",
       "2                                                                                                Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "3                                                                                                                                                                                                                     Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.   \n",
       "4                                                                                                                                                             Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .   \n",
       "...                                                                                                                                                                                                                                                                                                                                                  ...   \n",
       "195879                                                                                                                     And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!   \n",
       "195880  There's a Star Wars Expanded Universe comic featuring Luke's childhood friend Janek \"Tank\" Sunber, who'd joined the Empire, become a lieutenant, and ended up stationed on a planet of tribal aliens. The plot of that handful of comics is essentially Zulu, with Imperials desperately fighting wave after wave of aliens and being worn down.   \n",
       "195881                                                                                                                                                                                                      Private Thomas, one of the Welsh soldiers in Zulu, who gets chewed out for taking care of a calf instead of being on duty early in the film.   \n",
       "195882                                        According to Urban Legend, the Zulu warrior in Zulu who falls dying before his chief, and who makes a heartfelt declaration of loyalty as he does so, is actually saying something along the lines of \"I have an enormous penis\" in Swahilinotenote that Zulu people generally speak isiZulu, not Swahili.   \n",
       "195883                                                                                                                                                                                                                                     Zwartboek (Black Book): Deals with the Dutch resistance in the Nazi-occupied Netherlands during World War II.   \n",
       "\n",
       "       trope_id example_type  \n",
       "0        t22727          lit  \n",
       "1        t01294          lit  \n",
       "2        t22727          lit  \n",
       "3        t01080          lit  \n",
       "4        t25818          lit  \n",
       "...         ...          ...  \n",
       "195879   t12565         film  \n",
       "195880   t26776         film  \n",
       "195881   t08978         film  \n",
       "195882   t02282         film  \n",
       "195883   t12563         film  \n",
       "\n",
       "[195884 rows x 4 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trope_examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "1e2d9335-84c5-4625-b9a0-d82f78f3ed38",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8985"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trope_examples.Example.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "9014bfa6-efe4-496e-87cf-fe1cfb5fc310",
   "metadata": {},
   "outputs": [],
   "source": [
    "trope_examples = trope_examples.dropna(subset=['Example'])\n",
    "trope_examples = trope_examples.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "e31b3fbd-d9a9-4946-8dd3-f331d0fef3bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "trope_examples.to_csv('trope_examples.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "f59f2907-d9d1-47b2-b481-a9169f18a2fc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Example</th>\n",
       "      <th>trope_id</th>\n",
       "      <th>example_type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13</td>\n",
       "      <td>When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".</td>\n",
       "      <td>t01294</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15</td>\n",
       "      <td>Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16</td>\n",
       "      <td>Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.</td>\n",
       "      <td>t01080</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17</td>\n",
       "      <td>Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .</td>\n",
       "      <td>t25818</td>\n",
       "      <td>lit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195879</th>\n",
       "      <td>758972</td>\n",
       "      <td>And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!</td>\n",
       "      <td>t12565</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195880</th>\n",
       "      <td>758985</td>\n",
       "      <td>There's a Star Wars Expanded Universe comic featuring Luke's childhood friend Janek \"Tank\" Sunber, who'd joined the Empire, become a lieutenant, and ended up stationed on a planet of tribal aliens. The plot of that handful of comics is essentially Zulu, with Imperials desperately fighting wave after wave of aliens and being worn down.</td>\n",
       "      <td>t26776</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195881</th>\n",
       "      <td>759016</td>\n",
       "      <td>Private Thomas, one of the Welsh soldiers in Zulu, who gets chewed out for taking care of a calf instead of being on duty early in the film.</td>\n",
       "      <td>t08978</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195882</th>\n",
       "      <td>759046</td>\n",
       "      <td>According to Urban Legend, the Zulu warrior in Zulu who falls dying before his chief, and who makes a heartfelt declaration of loyalty as he does so, is actually saying something along the lines of \"I have an enormous penis\" in Swahilinotenote that Zulu people generally speak isiZulu, not Swahili.</td>\n",
       "      <td>t02282</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195883</th>\n",
       "      <td>759063</td>\n",
       "      <td>Zwartboek (Black Book): Deals with the Dutch resistance in the Nazi-occupied Netherlands during World War II.</td>\n",
       "      <td>t12563</td>\n",
       "      <td>film</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>186899 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         index  \\\n",
       "0           10   \n",
       "1           13   \n",
       "2           15   \n",
       "3           16   \n",
       "4           17   \n",
       "...        ...   \n",
       "195879  758972   \n",
       "195880  758985   \n",
       "195881  759016   \n",
       "195882  759046   \n",
       "195883  759063   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                                                                 Example  \\\n",
       "0                                                                            In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "1                                                                                                                                     When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".   \n",
       "2                                                                                                Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "3                                                                                                                                                                                                                     Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.   \n",
       "4                                                                                                                                                             Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .   \n",
       "...                                                                                                                                                                                                                                                                                                                                                  ...   \n",
       "195879                                                                                                                     And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!   \n",
       "195880  There's a Star Wars Expanded Universe comic featuring Luke's childhood friend Janek \"Tank\" Sunber, who'd joined the Empire, become a lieutenant, and ended up stationed on a planet of tribal aliens. The plot of that handful of comics is essentially Zulu, with Imperials desperately fighting wave after wave of aliens and being worn down.   \n",
       "195881                                                                                                                                                                                                      Private Thomas, one of the Welsh soldiers in Zulu, who gets chewed out for taking care of a calf instead of being on duty early in the film.   \n",
       "195882                                        According to Urban Legend, the Zulu warrior in Zulu who falls dying before his chief, and who makes a heartfelt declaration of loyalty as he does so, is actually saying something along the lines of \"I have an enormous penis\" in Swahilinotenote that Zulu people generally speak isiZulu, not Swahili.   \n",
       "195883                                                                                                                                                                                                                                     Zwartboek (Black Book): Deals with the Dutch resistance in the Nazi-occupied Netherlands during World War II.   \n",
       "\n",
       "       trope_id example_type  \n",
       "0        t22727          lit  \n",
       "1        t01294          lit  \n",
       "2        t22727          lit  \n",
       "3        t01080          lit  \n",
       "4        t25818          lit  \n",
       "...         ...          ...  \n",
       "195879   t12565         film  \n",
       "195880   t26776         film  \n",
       "195881   t08978         film  \n",
       "195882   t02282         film  \n",
       "195883   t12563         film  \n",
       "\n",
       "[186899 rows x 4 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trope_examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "bd77f911-9464-40bc-832e-b7be64793c98",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/json": {
       "ascii": false,
       "bar_format": null,
       "colour": null,
       "elapsed": 0.01701211929321289,
       "initial": 0,
       "n": 0,
       "ncols": null,
       "nrows": 29,
       "postfix": null,
       "prefix": "",
       "rate": null,
       "total": 288783,
       "unit": "it",
       "unit_divisor": 1000,
       "unit_scale": false
      },
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d30222cdc1e34931b2debda597a2c298",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/288783 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from get_vector import embed, get_similarity\n",
    "trope_examples = trope_examples.assign(embedding = trope_examples['Example'].progress_apply(embed))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bd120f1a-02c5-4abb-a5fd-ffd88e53ab80",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "2eecf9c8-c911-44ff-8bfd-1abf52145a2c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Example</th>\n",
       "      <th>trope_id</th>\n",
       "      <th>example_type</th>\n",
       "      <th>embedding</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "      <td>lit</td>\n",
       "      <td>(tf.Tensor(0.08800587, shape=(), dtype=float32), tf.Tensor(-0.08599387, shape=(), dtype=float32), tf.Tensor(-0.025380103, shape=(), dtype=float32), tf.Tensor(0.051938705, shape=(), dtype=float32), tf.Tensor(0.02031321, shape=(), dtype=float32), tf.Tensor(-0.0020598548, shape=(), dtype=float32), tf.Tensor(0.01301749, shape=(), dtype=float32), tf.Tensor(0.04751223, shape=(), dtype=float32), tf.Tensor(-0.0019549562, shape=(), dtype=float32), tf.Tensor(-0.05235701, shape=(), dtype=float32), tf.Tensor(-0.028665619, shape=(), dtype=float32), tf.Tensor(0.040233985, shape=(), dtype=float32), tf.Tensor(-0.0012896181, shape=(), dtype=float32), tf.Tensor(0.043469764, shape=(), dtype=float32), tf.Tensor(0.04371827, shape=(), dtype=float32), tf.Tensor(-0.012290332, shape=(), dtype=float32), tf.Tensor(0.08683354, shape=(), dtype=float32), tf.Tensor(0.021370953, shape=(), dtype=float32), tf.Tensor(-0.00017330215, shape=(), dtype=float32), tf.Tensor(-0.036185037, shape=(), dtype=float32), tf.Tensor(-0.04546309, shape=(), dtype=float32), tf.Tensor(0.037026417, shape=(), dtype=float32), tf.Tensor(0.058535624, shape=(), dtype=float32), tf.Tensor(-0.062196318, shape=(), dtype=float32), tf.Tensor(0.004359815, shape=(), dtype=float32), tf.Tensor(-0.002621128, shape=(), dtype=float32), tf.Tensor(-0.0023814337, shape=(), dtype=float32), tf.Tensor(0.079657964, shape=(), dtype=float32), tf.Tensor(0.016618242, shape=(), dtype=float32), tf.Tensor(-0.057619054, shape=(), dtype=float32), tf.Tensor(-0.06499728, shape=(), dtype=float32), tf.Tensor(-0.032943424, shape=(), dtype=float32), tf.Tensor(0.04166872, shape=(), dtype=float32), tf.Tensor(-0.0692408, shape=(), dtype=float32), tf.Tensor(0.004227797, shape=(), dtype=float32), tf.Tensor(-0.022902718, shape=(), dtype=float32), tf.Tensor(0.02059966, shape=(), dtype=float32), tf.Tensor(-0.00035657547, shape=(), dtype=float32), tf.Tensor(-0.029187722, shape=(), dtype=float32), tf.Tensor(-0.022358175, shape=(), dtype=float32), tf.Tensor(-0.022271976, shape=(), dtype=float32), tf.Tensor(0.04697008, shape=(), dtype=float32), tf.Tensor(-0.07002722, shape=(), dtype=float32), tf.Tensor(-0.027185066, shape=(), dtype=float32), tf.Tensor(-0.0030682483, shape=(), dtype=float32), tf.Tensor(-0.030895032, shape=(), dtype=float32), tf.Tensor(0.079647385, shape=(), dtype=float32), tf.Tensor(0.057742752, shape=(), dtype=float32), tf.Tensor(-0.025763107, shape=(), dtype=float32), tf.Tensor(0.07856635, shape=(), dtype=float32), tf.Tensor(-0.021441985, shape=(), dtype=float32), tf.Tensor(-0.0592378, shape=(), dtype=float32), tf.Tensor(0.00816241, shape=(), dtype=float32), tf.Tensor(-0.046447586, shape=(), dtype=float32), tf.Tensor(0.02299255, shape=(), dtype=float32), tf.Tensor(0.044731148, shape=(), dtype=float32), tf.Tensor(-0.053904887, shape=(), dtype=float32), tf.Tensor(0.008835485, shape=(), dtype=float32), tf.Tensor(0.031931505, shape=(), dtype=float32), tf.Tensor(0.00012375598, shape=(), dtype=float32), tf.Tensor(-0.009230669, shape=(), dtype=float32), tf.Tensor(-0.031224066, shape=(), dtype=float32), tf.Tensor(-0.02890986, shape=(), dtype=float32), tf.Tensor(0.053919934, shape=(), dtype=float32), tf.Tensor(-0.008776098, shape=(), dtype=float32), tf.Tensor(-0.030362194, shape=(), dtype=float32), tf.Tensor(-0.055678792, shape=(), dtype=float32), tf.Tensor(-0.017249918, shape=(), dtype=float32), tf.Tensor(-0.006067661, shape=(), dtype=float32), tf.Tensor(0.019166904, shape=(), dtype=float32), tf.Tensor(-0.023110569, shape=(), dtype=float32), tf.Tensor(-0.055081792, shape=(), dtype=float32), tf.Tensor(-0.0037445223, shape=(), dtype=float32), tf.Tensor(0.009316371, shape=(), dtype=float32), tf.Tensor(-0.06440608, shape=(), dtype=float32), tf.Tensor(-0.010236327, shape=(), dtype=float32), tf.Tensor(0.060167458, shape=(), dtype=float32), tf.Tensor(0.045644686, shape=(), dtype=float32), tf.Tensor(0.08203374, shape=(), dtype=float32), tf.Tensor(0.020977907, shape=(), dtype=float32), tf.Tensor(-0.07048969, shape=(), dtype=float32), tf.Tensor(-0.03973232, shape=(), dtype=float32), tf.Tensor(-0.009789303, shape=(), dtype=float32), tf.Tensor(-0.05202683, shape=(), dtype=float32), tf.Tensor(-0.054066382, shape=(), dtype=float32), tf.Tensor(-0.02701431, shape=(), dtype=float32), tf.Tensor(0.009901098, shape=(), dtype=float32), tf.Tensor(-0.02209809, shape=(), dtype=float32), tf.Tensor(0.08363872, shape=(), dtype=float32), tf.Tensor(0.008711273, shape=(), dtype=float32), tf.Tensor(-0.056222685, shape=(), dtype=float32), tf.Tensor(-0.022011213, shape=(), dtype=float32), tf.Tensor(0.053444725, shape=(), dtype=float32), tf.Tensor(0.01691654, shape=(), dtype=float32), tf.Tensor(-0.07186151, shape=(), dtype=float32), tf.Tensor(-0.05782236, shape=(), dtype=float32), tf.Tensor(0.007892808, shape=(), dtype=float32), tf.Tensor(-0.0030549834, shape=(), dtype=float32), tf.Tensor(-0.04067482, shape=(), dtype=float32), tf.Tensor(0.011997106, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".</td>\n",
       "      <td>t01294</td>\n",
       "      <td>lit</td>\n",
       "      <td>(tf.Tensor(0.010330548, shape=(), dtype=float32), tf.Tensor(-0.029361794, shape=(), dtype=float32), tf.Tensor(-0.02274592, shape=(), dtype=float32), tf.Tensor(-0.002653622, shape=(), dtype=float32), tf.Tensor(-0.030094229, shape=(), dtype=float32), tf.Tensor(0.07554659, shape=(), dtype=float32), tf.Tensor(0.0106902355, shape=(), dtype=float32), tf.Tensor(-0.010981429, shape=(), dtype=float32), tf.Tensor(0.026712697, shape=(), dtype=float32), tf.Tensor(0.026856365, shape=(), dtype=float32), tf.Tensor(-0.025904251, shape=(), dtype=float32), tf.Tensor(0.0047908076, shape=(), dtype=float32), tf.Tensor(0.0050791674, shape=(), dtype=float32), tf.Tensor(0.04752505, shape=(), dtype=float32), tf.Tensor(0.012012608, shape=(), dtype=float32), tf.Tensor(-0.004147278, shape=(), dtype=float32), tf.Tensor(0.130424, shape=(), dtype=float32), tf.Tensor(-0.01596583, shape=(), dtype=float32), tf.Tensor(0.007897065, shape=(), dtype=float32), tf.Tensor(-0.028737511, shape=(), dtype=float32), tf.Tensor(-0.060379915, shape=(), dtype=float32), tf.Tensor(0.047210198, shape=(), dtype=float32), tf.Tensor(0.062640674, shape=(), dtype=float32), tf.Tensor(0.019653106, shape=(), dtype=float32), tf.Tensor(-0.05946766, shape=(), dtype=float32), tf.Tensor(0.061379302, shape=(), dtype=float32), tf.Tensor(-0.08129871, shape=(), dtype=float32), tf.Tensor(0.0068943007, shape=(), dtype=float32), tf.Tensor(0.03735557, shape=(), dtype=float32), tf.Tensor(-0.04145611, shape=(), dtype=float32), tf.Tensor(-0.124526694, shape=(), dtype=float32), tf.Tensor(-0.0046156486, shape=(), dtype=float32), tf.Tensor(-0.019345704, shape=(), dtype=float32), tf.Tensor(0.04755795, shape=(), dtype=float32), tf.Tensor(-0.04352252, shape=(), dtype=float32), tf.Tensor(-0.023820508, shape=(), dtype=float32), tf.Tensor(-0.01023172, shape=(), dtype=float32), tf.Tensor(0.0007520435, shape=(), dtype=float32), tf.Tensor(-0.0076776952, shape=(), dtype=float32), tf.Tensor(-0.057130553, shape=(), dtype=float32), tf.Tensor(-0.0026168914, shape=(), dtype=float32), tf.Tensor(-0.023852542, shape=(), dtype=float32), tf.Tensor(-0.026527165, shape=(), dtype=float32), tf.Tensor(0.00721357, shape=(), dtype=float32), tf.Tensor(-0.048666157, shape=(), dtype=float32), tf.Tensor(0.017667599, shape=(), dtype=float32), tf.Tensor(0.05699326, shape=(), dtype=float32), tf.Tensor(0.010426503, shape=(), dtype=float32), tf.Tensor(0.01827146, shape=(), dtype=float32), tf.Tensor(0.02974568, shape=(), dtype=float32), tf.Tensor(-0.0019398523, shape=(), dtype=float32), tf.Tensor(-0.015875578, shape=(), dtype=float32), tf.Tensor(0.015058477, shape=(), dtype=float32), tf.Tensor(0.013088203, shape=(), dtype=float32), tf.Tensor(0.04991926, shape=(), dtype=float32), tf.Tensor(0.043303464, shape=(), dtype=float32), tf.Tensor(-0.015027079, shape=(), dtype=float32), tf.Tensor(-0.020142818, shape=(), dtype=float32), tf.Tensor(-0.01281049, shape=(), dtype=float32), tf.Tensor(0.07224833, shape=(), dtype=float32), tf.Tensor(-0.024062173, shape=(), dtype=float32), tf.Tensor(-0.054629564, shape=(), dtype=float32), tf.Tensor(-0.076675266, shape=(), dtype=float32), tf.Tensor(-0.031861987, shape=(), dtype=float32), tf.Tensor(0.035590075, shape=(), dtype=float32), tf.Tensor(0.043255467, shape=(), dtype=float32), tf.Tensor(-0.060752455, shape=(), dtype=float32), tf.Tensor(-0.03541548, shape=(), dtype=float32), tf.Tensor(0.03134576, shape=(), dtype=float32), tf.Tensor(0.0007455198, shape=(), dtype=float32), tf.Tensor(0.0042349533, shape=(), dtype=float32), tf.Tensor(0.010123341, shape=(), dtype=float32), tf.Tensor(0.059565395, shape=(), dtype=float32), tf.Tensor(-0.009693012, shape=(), dtype=float32), tf.Tensor(-0.017010922, shape=(), dtype=float32), tf.Tensor(-0.015895218, shape=(), dtype=float32), tf.Tensor(0.12877099, shape=(), dtype=float32), tf.Tensor(0.06468694, shape=(), dtype=float32), tf.Tensor(0.018707115, shape=(), dtype=float32), tf.Tensor(0.009936, shape=(), dtype=float32), tf.Tensor(-0.0043845777, shape=(), dtype=float32), tf.Tensor(-0.024331015, shape=(), dtype=float32), tf.Tensor(-0.036014624, shape=(), dtype=float32), tf.Tensor(-0.06364599, shape=(), dtype=float32), tf.Tensor(-0.048503984, shape=(), dtype=float32), tf.Tensor(-0.050774828, shape=(), dtype=float32), tf.Tensor(0.029971149, shape=(), dtype=float32), tf.Tensor(-0.024143403, shape=(), dtype=float32), tf.Tensor(0.06282789, shape=(), dtype=float32), tf.Tensor(-0.028649233, shape=(), dtype=float32), tf.Tensor(0.016530229, shape=(), dtype=float32), tf.Tensor(-0.019938834, shape=(), dtype=float32), tf.Tensor(0.06232069, shape=(), dtype=float32), tf.Tensor(0.045510154, shape=(), dtype=float32), tf.Tensor(-0.05315823, shape=(), dtype=float32), tf.Tensor(-0.09995444, shape=(), dtype=float32), tf.Tensor(-0.05620101, shape=(), dtype=float32), tf.Tensor(-0.01500156, shape=(), dtype=float32), tf.Tensor(-0.01709219, shape=(), dtype=float32), tf.Tensor(0.027486103, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.</td>\n",
       "      <td>t22727</td>\n",
       "      <td>lit</td>\n",
       "      <td>(tf.Tensor(0.07903765, shape=(), dtype=float32), tf.Tensor(-0.08122551, shape=(), dtype=float32), tf.Tensor(-0.029343614, shape=(), dtype=float32), tf.Tensor(0.03724364, shape=(), dtype=float32), tf.Tensor(0.011438731, shape=(), dtype=float32), tf.Tensor(-0.002793295, shape=(), dtype=float32), tf.Tensor(0.017981712, shape=(), dtype=float32), tf.Tensor(0.06524445, shape=(), dtype=float32), tf.Tensor(-0.018213881, shape=(), dtype=float32), tf.Tensor(-0.06824911, shape=(), dtype=float32), tf.Tensor(-0.03127585, shape=(), dtype=float32), tf.Tensor(0.06952839, shape=(), dtype=float32), tf.Tensor(-0.017614137, shape=(), dtype=float32), tf.Tensor(0.055293288, shape=(), dtype=float32), tf.Tensor(0.040941577, shape=(), dtype=float32), tf.Tensor(-0.023553396, shape=(), dtype=float32), tf.Tensor(0.09257069, shape=(), dtype=float32), tf.Tensor(0.022809014, shape=(), dtype=float32), tf.Tensor(0.007699889, shape=(), dtype=float32), tf.Tensor(-0.012788549, shape=(), dtype=float32), tf.Tensor(-0.04757676, shape=(), dtype=float32), tf.Tensor(0.028746195, shape=(), dtype=float32), tf.Tensor(0.05641736, shape=(), dtype=float32), tf.Tensor(-0.054203033, shape=(), dtype=float32), tf.Tensor(-0.010500583, shape=(), dtype=float32), tf.Tensor(-0.004929476, shape=(), dtype=float32), tf.Tensor(-0.006849469, shape=(), dtype=float32), tf.Tensor(0.061456073, shape=(), dtype=float32), tf.Tensor(0.023601279, shape=(), dtype=float32), tf.Tensor(-0.05851914, shape=(), dtype=float32), tf.Tensor(-0.063357376, shape=(), dtype=float32), tf.Tensor(-0.050873302, shape=(), dtype=float32), tf.Tensor(0.053865567, shape=(), dtype=float32), tf.Tensor(-0.059307277, shape=(), dtype=float32), tf.Tensor(0.0042104274, shape=(), dtype=float32), tf.Tensor(-0.035044774, shape=(), dtype=float32), tf.Tensor(0.013260701, shape=(), dtype=float32), tf.Tensor(-0.0073710675, shape=(), dtype=float32), tf.Tensor(-0.022457961, shape=(), dtype=float32), tf.Tensor(-0.02190834, shape=(), dtype=float32), tf.Tensor(-0.02646201, shape=(), dtype=float32), tf.Tensor(0.05482588, shape=(), dtype=float32), tf.Tensor(-0.050045248, shape=(), dtype=float32), tf.Tensor(-0.026658569, shape=(), dtype=float32), tf.Tensor(-0.0026801967, shape=(), dtype=float32), tf.Tensor(-0.022143172, shape=(), dtype=float32), tf.Tensor(0.08452685, shape=(), dtype=float32), tf.Tensor(0.052563205, shape=(), dtype=float32), tf.Tensor(-0.028089037, shape=(), dtype=float32), tf.Tensor(0.048151925, shape=(), dtype=float32), tf.Tensor(-0.01412995, shape=(), dtype=float32), tf.Tensor(-0.04600642, shape=(), dtype=float32), tf.Tensor(0.014064551, shape=(), dtype=float32), tf.Tensor(-0.06108639, shape=(), dtype=float32), tf.Tensor(0.014461201, shape=(), dtype=float32), tf.Tensor(0.039829098, shape=(), dtype=float32), tf.Tensor(-0.038568724, shape=(), dtype=float32), tf.Tensor(0.003268224, shape=(), dtype=float32), tf.Tensor(0.029916162, shape=(), dtype=float32), tf.Tensor(-0.011676105, shape=(), dtype=float32), tf.Tensor(-0.013288738, shape=(), dtype=float32), tf.Tensor(-0.037835382, shape=(), dtype=float32), tf.Tensor(-0.048727572, shape=(), dtype=float32), tf.Tensor(0.063617, shape=(), dtype=float32), tf.Tensor(-0.018076634, shape=(), dtype=float32), tf.Tensor(-0.01990476, shape=(), dtype=float32), tf.Tensor(-0.042753935, shape=(), dtype=float32), tf.Tensor(-0.018533094, shape=(), dtype=float32), tf.Tensor(-0.00074974366, shape=(), dtype=float32), tf.Tensor(0.0008349011, shape=(), dtype=float32), tf.Tensor(-0.012258501, shape=(), dtype=float32), tf.Tensor(-0.02587158, shape=(), dtype=float32), tf.Tensor(-0.0052914466, shape=(), dtype=float32), tf.Tensor(0.00012988648, shape=(), dtype=float32), tf.Tensor(-0.051806636, shape=(), dtype=float32), tf.Tensor(-0.0068446877, shape=(), dtype=float32), tf.Tensor(0.045981146, shape=(), dtype=float32), tf.Tensor(0.014970127, shape=(), dtype=float32), tf.Tensor(0.07396664, shape=(), dtype=float32), tf.Tensor(0.0449271, shape=(), dtype=float32), tf.Tensor(-0.05091219, shape=(), dtype=float32), tf.Tensor(-0.03671727, shape=(), dtype=float32), tf.Tensor(-0.027304694, shape=(), dtype=float32), tf.Tensor(-0.04140609, shape=(), dtype=float32), tf.Tensor(-0.0636552, shape=(), dtype=float32), tf.Tensor(-0.032051597, shape=(), dtype=float32), tf.Tensor(0.023820529, shape=(), dtype=float32), tf.Tensor(-0.0074291355, shape=(), dtype=float32), tf.Tensor(0.09682213, shape=(), dtype=float32), tf.Tensor(0.020731114, shape=(), dtype=float32), tf.Tensor(-0.07162586, shape=(), dtype=float32), tf.Tensor(-0.017131131, shape=(), dtype=float32), tf.Tensor(0.05679401, shape=(), dtype=float32), tf.Tensor(0.020841591, shape=(), dtype=float32), tf.Tensor(-0.07075116, shape=(), dtype=float32), tf.Tensor(-0.06193501, shape=(), dtype=float32), tf.Tensor(0.006631097, shape=(), dtype=float32), tf.Tensor(-0.0039047645, shape=(), dtype=float32), tf.Tensor(-0.04713479, shape=(), dtype=float32), tf.Tensor(-0.015411815, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.</td>\n",
       "      <td>t01080</td>\n",
       "      <td>lit</td>\n",
       "      <td>(tf.Tensor(-0.03549712, shape=(), dtype=float32), tf.Tensor(0.03641831, shape=(), dtype=float32), tf.Tensor(0.018138606, shape=(), dtype=float32), tf.Tensor(0.04770905, shape=(), dtype=float32), tf.Tensor(-0.004124664, shape=(), dtype=float32), tf.Tensor(0.0024107983, shape=(), dtype=float32), tf.Tensor(-0.040756386, shape=(), dtype=float32), tf.Tensor(-0.034914747, shape=(), dtype=float32), tf.Tensor(0.020394739, shape=(), dtype=float32), tf.Tensor(0.0172372, shape=(), dtype=float32), tf.Tensor(-0.041197516, shape=(), dtype=float32), tf.Tensor(-0.026053572, shape=(), dtype=float32), tf.Tensor(-0.02246861, shape=(), dtype=float32), tf.Tensor(0.017176801, shape=(), dtype=float32), tf.Tensor(0.032556154, shape=(), dtype=float32), tf.Tensor(0.021616759, shape=(), dtype=float32), tf.Tensor(0.136648, shape=(), dtype=float32), tf.Tensor(0.05070697, shape=(), dtype=float32), tf.Tensor(-0.0370852, shape=(), dtype=float32), tf.Tensor(0.028701453, shape=(), dtype=float32), tf.Tensor(-0.059691045, shape=(), dtype=float32), tf.Tensor(0.010044211, shape=(), dtype=float32), tf.Tensor(0.059767265, shape=(), dtype=float32), tf.Tensor(0.011906863, shape=(), dtype=float32), tf.Tensor(-0.00084792287, shape=(), dtype=float32), tf.Tensor(0.01737797, shape=(), dtype=float32), tf.Tensor(-0.011986313, shape=(), dtype=float32), tf.Tensor(0.10150425, shape=(), dtype=float32), tf.Tensor(0.07557448, shape=(), dtype=float32), tf.Tensor(-0.037511464, shape=(), dtype=float32), tf.Tensor(-0.0392109, shape=(), dtype=float32), tf.Tensor(-0.026490528, shape=(), dtype=float32), tf.Tensor(-0.05747621, shape=(), dtype=float32), tf.Tensor(0.058619354, shape=(), dtype=float32), tf.Tensor(-0.06631421, shape=(), dtype=float32), tf.Tensor(-0.051940884, shape=(), dtype=float32), tf.Tensor(0.02359717, shape=(), dtype=float32), tf.Tensor(0.023845933, shape=(), dtype=float32), tf.Tensor(0.055003658, shape=(), dtype=float32), tf.Tensor(-0.0891166, shape=(), dtype=float32), tf.Tensor(0.047206674, shape=(), dtype=float32), tf.Tensor(-0.04001871, shape=(), dtype=float32), tf.Tensor(0.046865817, shape=(), dtype=float32), tf.Tensor(0.003833754, shape=(), dtype=float32), tf.Tensor(-0.050639827, shape=(), dtype=float32), tf.Tensor(-0.009351717, shape=(), dtype=float32), tf.Tensor(0.06790367, shape=(), dtype=float32), tf.Tensor(0.028699124, shape=(), dtype=float32), tf.Tensor(0.036173567, shape=(), dtype=float32), tf.Tensor(-0.017418522, shape=(), dtype=float32), tf.Tensor(0.00669771, shape=(), dtype=float32), tf.Tensor(-0.05520239, shape=(), dtype=float32), tf.Tensor(-0.037526235, shape=(), dtype=float32), tf.Tensor(0.0018540716, shape=(), dtype=float32), tf.Tensor(-0.032877833, shape=(), dtype=float32), tf.Tensor(-0.0038160027, shape=(), dtype=float32), tf.Tensor(0.04226888, shape=(), dtype=float32), tf.Tensor(-0.010834188, shape=(), dtype=float32), tf.Tensor(0.06358276, shape=(), dtype=float32), tf.Tensor(-0.016166119, shape=(), dtype=float32), tf.Tensor(-0.013715793, shape=(), dtype=float32), tf.Tensor(-0.024011398, shape=(), dtype=float32), tf.Tensor(-0.0028317412, shape=(), dtype=float32), tf.Tensor(-0.012698178, shape=(), dtype=float32), tf.Tensor(-0.06302096, shape=(), dtype=float32), tf.Tensor(0.0494479, shape=(), dtype=float32), tf.Tensor(-0.01340376, shape=(), dtype=float32), tf.Tensor(-0.023631789, shape=(), dtype=float32), tf.Tensor(0.09890174, shape=(), dtype=float32), tf.Tensor(-0.08014861, shape=(), dtype=float32), tf.Tensor(0.06290152, shape=(), dtype=float32), tf.Tensor(0.031284846, shape=(), dtype=float32), tf.Tensor(-0.06275099, shape=(), dtype=float32), tf.Tensor(0.041656412, shape=(), dtype=float32), tf.Tensor(0.036877573, shape=(), dtype=float32), tf.Tensor(0.033646654, shape=(), dtype=float32), tf.Tensor(0.039049953, shape=(), dtype=float32), tf.Tensor(-0.059652343, shape=(), dtype=float32), tf.Tensor(0.03144189, shape=(), dtype=float32), tf.Tensor(-0.016832195, shape=(), dtype=float32), tf.Tensor(-0.05423973, shape=(), dtype=float32), tf.Tensor(-0.046023183, shape=(), dtype=float32), tf.Tensor(-0.075923145, shape=(), dtype=float32), tf.Tensor(-0.03515708, shape=(), dtype=float32), tf.Tensor(-0.020270295, shape=(), dtype=float32), tf.Tensor(0.0070862304, shape=(), dtype=float32), tf.Tensor(0.014913817, shape=(), dtype=float32), tf.Tensor(0.059813924, shape=(), dtype=float32), tf.Tensor(0.020985503, shape=(), dtype=float32), tf.Tensor(-0.026704855, shape=(), dtype=float32), tf.Tensor(-0.09044281, shape=(), dtype=float32), tf.Tensor(0.0038491231, shape=(), dtype=float32), tf.Tensor(0.046002652, shape=(), dtype=float32), tf.Tensor(0.07362481, shape=(), dtype=float32), tf.Tensor(-0.058121517, shape=(), dtype=float32), tf.Tensor(-0.034743812, shape=(), dtype=float32), tf.Tensor(0.03586781, shape=(), dtype=float32), tf.Tensor(-0.02638692, shape=(), dtype=float32), tf.Tensor(-0.07126451, shape=(), dtype=float32), tf.Tensor(-0.012174028, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .</td>\n",
       "      <td>t25818</td>\n",
       "      <td>lit</td>\n",
       "      <td>(tf.Tensor(0.023967337, shape=(), dtype=float32), tf.Tensor(-0.084397495, shape=(), dtype=float32), tf.Tensor(0.01415953, shape=(), dtype=float32), tf.Tensor(-0.01036743, shape=(), dtype=float32), tf.Tensor(0.035001464, shape=(), dtype=float32), tf.Tensor(0.005071171, shape=(), dtype=float32), tf.Tensor(-0.01694777, shape=(), dtype=float32), tf.Tensor(-0.03718967, shape=(), dtype=float32), tf.Tensor(0.024154918, shape=(), dtype=float32), tf.Tensor(-0.050603937, shape=(), dtype=float32), tf.Tensor(-0.005885336, shape=(), dtype=float32), tf.Tensor(0.07657696, shape=(), dtype=float32), tf.Tensor(-0.05108257, shape=(), dtype=float32), tf.Tensor(0.03954993, shape=(), dtype=float32), tf.Tensor(0.057510085, shape=(), dtype=float32), tf.Tensor(-0.018536614, shape=(), dtype=float32), tf.Tensor(0.12386522, shape=(), dtype=float32), tf.Tensor(0.032914054, shape=(), dtype=float32), tf.Tensor(-0.006091986, shape=(), dtype=float32), tf.Tensor(-0.0071318448, shape=(), dtype=float32), tf.Tensor(-0.07268326, shape=(), dtype=float32), tf.Tensor(-0.01380359, shape=(), dtype=float32), tf.Tensor(0.018614434, shape=(), dtype=float32), tf.Tensor(-0.0332584, shape=(), dtype=float32), tf.Tensor(-0.026341196, shape=(), dtype=float32), tf.Tensor(-0.021997431, shape=(), dtype=float32), tf.Tensor(0.0015622203, shape=(), dtype=float32), tf.Tensor(0.044560604, shape=(), dtype=float32), tf.Tensor(-0.015563141, shape=(), dtype=float32), tf.Tensor(-0.05528924, shape=(), dtype=float32), tf.Tensor(-0.069653936, shape=(), dtype=float32), tf.Tensor(-0.037985787, shape=(), dtype=float32), tf.Tensor(0.066987574, shape=(), dtype=float32), tf.Tensor(0.0042458903, shape=(), dtype=float32), tf.Tensor(-0.0039066095, shape=(), dtype=float32), tf.Tensor(0.00080658443, shape=(), dtype=float32), tf.Tensor(-0.018237745, shape=(), dtype=float32), tf.Tensor(0.013343024, shape=(), dtype=float32), tf.Tensor(-0.031679623, shape=(), dtype=float32), tf.Tensor(-0.021655528, shape=(), dtype=float32), tf.Tensor(0.018567353, shape=(), dtype=float32), tf.Tensor(0.055041708, shape=(), dtype=float32), tf.Tensor(-0.064987734, shape=(), dtype=float32), tf.Tensor(-0.0026901325, shape=(), dtype=float32), tf.Tensor(-0.029532868, shape=(), dtype=float32), tf.Tensor(-0.043519426, shape=(), dtype=float32), tf.Tensor(0.048361097, shape=(), dtype=float32), tf.Tensor(0.026961464, shape=(), dtype=float32), tf.Tensor(0.022430457, shape=(), dtype=float32), tf.Tensor(0.054570664, shape=(), dtype=float32), tf.Tensor(0.0007716413, shape=(), dtype=float32), tf.Tensor(-0.023416478, shape=(), dtype=float32), tf.Tensor(0.036561195, shape=(), dtype=float32), tf.Tensor(-0.05937416, shape=(), dtype=float32), tf.Tensor(-0.045613628, shape=(), dtype=float32), tf.Tensor(0.060605895, shape=(), dtype=float32), tf.Tensor(0.025672719, shape=(), dtype=float32), tf.Tensor(-0.00018210869, shape=(), dtype=float32), tf.Tensor(0.04163569, shape=(), dtype=float32), tf.Tensor(0.029721607, shape=(), dtype=float32), tf.Tensor(0.005982857, shape=(), dtype=float32), tf.Tensor(-0.033309005, shape=(), dtype=float32), tf.Tensor(-0.05784923, shape=(), dtype=float32), tf.Tensor(0.039723407, shape=(), dtype=float32), tf.Tensor(-0.04391734, shape=(), dtype=float32), tf.Tensor(0.0036102408, shape=(), dtype=float32), tf.Tensor(-0.052760348, shape=(), dtype=float32), tf.Tensor(-0.021783441, shape=(), dtype=float32), tf.Tensor(-0.004423645, shape=(), dtype=float32), tf.Tensor(-0.011782224, shape=(), dtype=float32), tf.Tensor(0.0259689, shape=(), dtype=float32), tf.Tensor(-0.020055423, shape=(), dtype=float32), tf.Tensor(-0.01234067, shape=(), dtype=float32), tf.Tensor(0.020949947, shape=(), dtype=float32), tf.Tensor(-0.029320963, shape=(), dtype=float32), tf.Tensor(0.009272562, shape=(), dtype=float32), tf.Tensor(-0.004329086, shape=(), dtype=float32), tf.Tensor(-0.017080763, shape=(), dtype=float32), tf.Tensor(0.057862487, shape=(), dtype=float32), tf.Tensor(0.04540068, shape=(), dtype=float32), tf.Tensor(-0.048767716, shape=(), dtype=float32), tf.Tensor(-0.014069336, shape=(), dtype=float32), tf.Tensor(-0.018453393, shape=(), dtype=float32), tf.Tensor(-0.045883752, shape=(), dtype=float32), tf.Tensor(-0.053496797, shape=(), dtype=float32), tf.Tensor(0.03852861, shape=(), dtype=float32), tf.Tensor(0.03908949, shape=(), dtype=float32), tf.Tensor(-0.013852586, shape=(), dtype=float32), tf.Tensor(0.03313901, shape=(), dtype=float32), tf.Tensor(0.026762994, shape=(), dtype=float32), tf.Tensor(-0.103191234, shape=(), dtype=float32), tf.Tensor(-0.021607641, shape=(), dtype=float32), tf.Tensor(0.035765253, shape=(), dtype=float32), tf.Tensor(0.03716235, shape=(), dtype=float32), tf.Tensor(-0.040432636, shape=(), dtype=float32), tf.Tensor(0.0053194156, shape=(), dtype=float32), tf.Tensor(0.020130217, shape=(), dtype=float32), tf.Tensor(-0.011448691, shape=(), dtype=float32), tf.Tensor(-0.030330446, shape=(), dtype=float32), tf.Tensor(-0.039985277, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300079</th>\n",
       "      <td>All the caballeros are California placenames with \"Don\" instead of \"San\"  Diego, Francisco, Fernando, Jose,  Luis Obispo .</td>\n",
       "      <td>t23159</td>\n",
       "      <td>film</td>\n",
       "      <td>(tf.Tensor(-0.056905918, shape=(), dtype=float32), tf.Tensor(-0.029011853, shape=(), dtype=float32), tf.Tensor(-0.032758486, shape=(), dtype=float32), tf.Tensor(-0.033120498, shape=(), dtype=float32), tf.Tensor(-0.016638504, shape=(), dtype=float32), tf.Tensor(-0.03368108, shape=(), dtype=float32), tf.Tensor(0.028499708, shape=(), dtype=float32), tf.Tensor(-0.01722193, shape=(), dtype=float32), tf.Tensor(-0.059155736, shape=(), dtype=float32), tf.Tensor(0.06507286, shape=(), dtype=float32), tf.Tensor(0.02724566, shape=(), dtype=float32), tf.Tensor(-0.0013793716, shape=(), dtype=float32), tf.Tensor(-0.103186354, shape=(), dtype=float32), tf.Tensor(0.083762184, shape=(), dtype=float32), tf.Tensor(0.04576251, shape=(), dtype=float32), tf.Tensor(-0.058948424, shape=(), dtype=float32), tf.Tensor(0.074407704, shape=(), dtype=float32), tf.Tensor(-0.0143182445, shape=(), dtype=float32), tf.Tensor(0.03878837, shape=(), dtype=float32), tf.Tensor(0.074402735, shape=(), dtype=float32), tf.Tensor(0.021511186, shape=(), dtype=float32), tf.Tensor(0.01120709, shape=(), dtype=float32), tf.Tensor(-0.0055965506, shape=(), dtype=float32), tf.Tensor(-0.03297675, shape=(), dtype=float32), tf.Tensor(-0.017122181, shape=(), dtype=float32), tf.Tensor(0.023215504, shape=(), dtype=float32), tf.Tensor(0.0023373598, shape=(), dtype=float32), tf.Tensor(-0.020384455, shape=(), dtype=float32), tf.Tensor(-0.017145824, shape=(), dtype=float32), tf.Tensor(-0.04054259, shape=(), dtype=float32), tf.Tensor(0.0015166264, shape=(), dtype=float32), tf.Tensor(0.009659941, shape=(), dtype=float32), tf.Tensor(-0.011240177, shape=(), dtype=float32), tf.Tensor(-0.13036893, shape=(), dtype=float32), tf.Tensor(0.016040368, shape=(), dtype=float32), tf.Tensor(-0.017487042, shape=(), dtype=float32), tf.Tensor(-0.00030466286, shape=(), dtype=float32), tf.Tensor(0.05042587, shape=(), dtype=float32), tf.Tensor(0.021934887, shape=(), dtype=float32), tf.Tensor(0.021242432, shape=(), dtype=float32), tf.Tensor(0.023614423, shape=(), dtype=float32), tf.Tensor(0.11073799, shape=(), dtype=float32), tf.Tensor(-0.035394892, shape=(), dtype=float32), tf.Tensor(0.03901027, shape=(), dtype=float32), tf.Tensor(0.007175722, shape=(), dtype=float32), tf.Tensor(0.013500772, shape=(), dtype=float32), tf.Tensor(-0.06808823, shape=(), dtype=float32), tf.Tensor(-0.016812729, shape=(), dtype=float32), tf.Tensor(-0.02199846, shape=(), dtype=float32), tf.Tensor(-0.034011316, shape=(), dtype=float32), tf.Tensor(0.006890839, shape=(), dtype=float32), tf.Tensor(-0.0077404124, shape=(), dtype=float32), tf.Tensor(-0.016576642, shape=(), dtype=float32), tf.Tensor(-0.011995097, shape=(), dtype=float32), tf.Tensor(-0.03620341, shape=(), dtype=float32), tf.Tensor(0.050961435, shape=(), dtype=float32), tf.Tensor(-4.2990498e-05, shape=(), dtype=float32), tf.Tensor(0.034567915, shape=(), dtype=float32), tf.Tensor(-0.00025924182, shape=(), dtype=float32), tf.Tensor(0.016785251, shape=(), dtype=float32), tf.Tensor(0.0970638, shape=(), dtype=float32), tf.Tensor(0.042462207, shape=(), dtype=float32), tf.Tensor(-0.0085408, shape=(), dtype=float32), tf.Tensor(0.0313212, shape=(), dtype=float32), tf.Tensor(-0.009317408, shape=(), dtype=float32), tf.Tensor(0.06608054, shape=(), dtype=float32), tf.Tensor(-0.009573651, shape=(), dtype=float32), tf.Tensor(-0.002945682, shape=(), dtype=float32), tf.Tensor(-0.023405321, shape=(), dtype=float32), tf.Tensor(-0.03322177, shape=(), dtype=float32), tf.Tensor(-0.0069239573, shape=(), dtype=float32), tf.Tensor(-0.04210521, shape=(), dtype=float32), tf.Tensor(0.011191868, shape=(), dtype=float32), tf.Tensor(-0.04007502, shape=(), dtype=float32), tf.Tensor(-0.041769423, shape=(), dtype=float32), tf.Tensor(-0.036197573, shape=(), dtype=float32), tf.Tensor(0.005983288, shape=(), dtype=float32), tf.Tensor(-0.056538347, shape=(), dtype=float32), tf.Tensor(-0.065433554, shape=(), dtype=float32), tf.Tensor(0.0010896284, shape=(), dtype=float32), tf.Tensor(0.016883992, shape=(), dtype=float32), tf.Tensor(0.010104156, shape=(), dtype=float32), tf.Tensor(-0.017128741, shape=(), dtype=float32), tf.Tensor(-0.06581602, shape=(), dtype=float32), tf.Tensor(0.040816627, shape=(), dtype=float32), tf.Tensor(0.0073410994, shape=(), dtype=float32), tf.Tensor(-0.0074424613, shape=(), dtype=float32), tf.Tensor(-0.057583623, shape=(), dtype=float32), tf.Tensor(0.006783987, shape=(), dtype=float32), tf.Tensor(-0.020749316, shape=(), dtype=float32), tf.Tensor(-0.01439232, shape=(), dtype=float32), tf.Tensor(-0.0389859, shape=(), dtype=float32), tf.Tensor(0.10855435, shape=(), dtype=float32), tf.Tensor(0.052365422, shape=(), dtype=float32), tf.Tensor(-0.055226468, shape=(), dtype=float32), tf.Tensor(0.015788274, shape=(), dtype=float32), tf.Tensor(-0.04391703, shape=(), dtype=float32), tf.Tensor(-0.05178745, shape=(), dtype=float32), tf.Tensor(0.043569602, shape=(), dtype=float32), tf.Tensor(-0.07885928, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300080</th>\n",
       "      <td>\"Sink me!\"</td>\n",
       "      <td>t03382</td>\n",
       "      <td>film</td>\n",
       "      <td>(tf.Tensor(0.06873965, shape=(), dtype=float32), tf.Tensor(0.011879233, shape=(), dtype=float32), tf.Tensor(0.009017263, shape=(), dtype=float32), tf.Tensor(-0.018235395, shape=(), dtype=float32), tf.Tensor(-0.07004947, shape=(), dtype=float32), tf.Tensor(-0.02003451, shape=(), dtype=float32), tf.Tensor(0.07945702, shape=(), dtype=float32), tf.Tensor(-0.048830975, shape=(), dtype=float32), tf.Tensor(-0.024501288, shape=(), dtype=float32), tf.Tensor(-0.02304068, shape=(), dtype=float32), tf.Tensor(-0.003356253, shape=(), dtype=float32), tf.Tensor(0.06627553, shape=(), dtype=float32), tf.Tensor(0.011041261, shape=(), dtype=float32), tf.Tensor(0.04809818, shape=(), dtype=float32), tf.Tensor(-0.09761175, shape=(), dtype=float32), tf.Tensor(-0.04054614, shape=(), dtype=float32), tf.Tensor(0.059656307, shape=(), dtype=float32), tf.Tensor(-0.059224084, shape=(), dtype=float32), tf.Tensor(0.026933756, shape=(), dtype=float32), tf.Tensor(-0.026851019, shape=(), dtype=float32), tf.Tensor(0.026299762, shape=(), dtype=float32), tf.Tensor(0.0060137818, shape=(), dtype=float32), tf.Tensor(0.011821149, shape=(), dtype=float32), tf.Tensor(0.041414645, shape=(), dtype=float32), tf.Tensor(-0.025367305, shape=(), dtype=float32), tf.Tensor(-0.006416064, shape=(), dtype=float32), tf.Tensor(-0.010305901, shape=(), dtype=float32), tf.Tensor(-0.004629364, shape=(), dtype=float32), tf.Tensor(0.006930405, shape=(), dtype=float32), tf.Tensor(0.090073176, shape=(), dtype=float32), tf.Tensor(0.03630198, shape=(), dtype=float32), tf.Tensor(0.028755702, shape=(), dtype=float32), tf.Tensor(-0.007877648, shape=(), dtype=float32), tf.Tensor(0.025796896, shape=(), dtype=float32), tf.Tensor(-0.062035393, shape=(), dtype=float32), tf.Tensor(-0.0046287896, shape=(), dtype=float32), tf.Tensor(0.014067956, shape=(), dtype=float32), tf.Tensor(-0.012977073, shape=(), dtype=float32), tf.Tensor(0.0029820544, shape=(), dtype=float32), tf.Tensor(-0.09998014, shape=(), dtype=float32), tf.Tensor(-0.045328774, shape=(), dtype=float32), tf.Tensor(0.0052359262, shape=(), dtype=float32), tf.Tensor(-0.044200826, shape=(), dtype=float32), tf.Tensor(-0.03938994, shape=(), dtype=float32), tf.Tensor(0.04806144, shape=(), dtype=float32), tf.Tensor(-0.014352055, shape=(), dtype=float32), tf.Tensor(-0.022317957, shape=(), dtype=float32), tf.Tensor(0.071658514, shape=(), dtype=float32), tf.Tensor(0.03830682, shape=(), dtype=float32), tf.Tensor(-0.03129089, shape=(), dtype=float32), tf.Tensor(-0.03567564, shape=(), dtype=float32), tf.Tensor(0.03654407, shape=(), dtype=float32), tf.Tensor(0.04357404, shape=(), dtype=float32), tf.Tensor(-0.027244585, shape=(), dtype=float32), tf.Tensor(0.04282825, shape=(), dtype=float32), tf.Tensor(-0.013412595, shape=(), dtype=float32), tf.Tensor(0.038874693, shape=(), dtype=float32), tf.Tensor(0.05319179, shape=(), dtype=float32), tf.Tensor(-0.0023268433, shape=(), dtype=float32), tf.Tensor(0.01715461, shape=(), dtype=float32), tf.Tensor(-0.013560631, shape=(), dtype=float32), tf.Tensor(-0.040419366, shape=(), dtype=float32), tf.Tensor(-0.006181546, shape=(), dtype=float32), tf.Tensor(0.042837176, shape=(), dtype=float32), tf.Tensor(-0.0055621546, shape=(), dtype=float32), tf.Tensor(0.0043764277, shape=(), dtype=float32), tf.Tensor(0.009759368, shape=(), dtype=float32), tf.Tensor(-0.015365781, shape=(), dtype=float32), tf.Tensor(-0.019109767, shape=(), dtype=float32), tf.Tensor(-0.06670105, shape=(), dtype=float32), tf.Tensor(-0.058795642, shape=(), dtype=float32), tf.Tensor(0.029145777, shape=(), dtype=float32), tf.Tensor(0.020474376, shape=(), dtype=float32), tf.Tensor(-0.0052547404, shape=(), dtype=float32), tf.Tensor(-0.010090535, shape=(), dtype=float32), tf.Tensor(0.07101111, shape=(), dtype=float32), tf.Tensor(0.045605715, shape=(), dtype=float32), tf.Tensor(-0.02296545, shape=(), dtype=float32), tf.Tensor(-0.021467282, shape=(), dtype=float32), tf.Tensor(-0.071915306, shape=(), dtype=float32), tf.Tensor(0.008871779, shape=(), dtype=float32), tf.Tensor(0.01678863, shape=(), dtype=float32), tf.Tensor(-0.016889704, shape=(), dtype=float32), tf.Tensor(0.039395362, shape=(), dtype=float32), tf.Tensor(-0.035669666, shape=(), dtype=float32), tf.Tensor(0.005358715, shape=(), dtype=float32), tf.Tensor(0.016329113, shape=(), dtype=float32), tf.Tensor(-0.026122803, shape=(), dtype=float32), tf.Tensor(0.026374765, shape=(), dtype=float32), tf.Tensor(0.01715521, shape=(), dtype=float32), tf.Tensor(0.008343227, shape=(), dtype=float32), tf.Tensor(-0.014655218, shape=(), dtype=float32), tf.Tensor(-0.019896802, shape=(), dtype=float32), tf.Tensor(0.09332796, shape=(), dtype=float32), tf.Tensor(-0.0077383136, shape=(), dtype=float32), tf.Tensor(-0.020521123, shape=(), dtype=float32), tf.Tensor(-8.6800566e-05, shape=(), dtype=float32), tf.Tensor(0.0622379, shape=(), dtype=float32), tf.Tensor(0.039992284, shape=(), dtype=float32), tf.Tensor(-0.067760274, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300081</th>\n",
       "      <td>Diego jumps from a second-floor window to escape pursuit ... and injures himself severely enough to force his Zorroing into a siesta.</td>\n",
       "      <td>t18297</td>\n",
       "      <td>film</td>\n",
       "      <td>(tf.Tensor(0.023907334, shape=(), dtype=float32), tf.Tensor(0.052315816, shape=(), dtype=float32), tf.Tensor(-0.051633414, shape=(), dtype=float32), tf.Tensor(0.0077471714, shape=(), dtype=float32), tf.Tensor(0.03970935, shape=(), dtype=float32), tf.Tensor(-0.013736675, shape=(), dtype=float32), tf.Tensor(-0.012582323, shape=(), dtype=float32), tf.Tensor(-0.024585912, shape=(), dtype=float32), tf.Tensor(-0.014852009, shape=(), dtype=float32), tf.Tensor(-0.055619143, shape=(), dtype=float32), tf.Tensor(0.008263253, shape=(), dtype=float32), tf.Tensor(-0.0047130096, shape=(), dtype=float32), tf.Tensor(-0.032742828, shape=(), dtype=float32), tf.Tensor(0.05743263, shape=(), dtype=float32), tf.Tensor(-0.01613141, shape=(), dtype=float32), tf.Tensor(-0.034406934, shape=(), dtype=float32), tf.Tensor(0.038817033, shape=(), dtype=float32), tf.Tensor(0.019072218, shape=(), dtype=float32), tf.Tensor(0.011669539, shape=(), dtype=float32), tf.Tensor(-0.090077706, shape=(), dtype=float32), tf.Tensor(0.0014555468, shape=(), dtype=float32), tf.Tensor(0.005513809, shape=(), dtype=float32), tf.Tensor(-0.030070733, shape=(), dtype=float32), tf.Tensor(-0.010374811, shape=(), dtype=float32), tf.Tensor(0.039355654, shape=(), dtype=float32), tf.Tensor(-0.037163842, shape=(), dtype=float32), tf.Tensor(0.02772324, shape=(), dtype=float32), tf.Tensor(-0.02353395, shape=(), dtype=float32), tf.Tensor(0.028946167, shape=(), dtype=float32), tf.Tensor(0.017719546, shape=(), dtype=float32), tf.Tensor(0.0018260363, shape=(), dtype=float32), tf.Tensor(0.00028033735, shape=(), dtype=float32), tf.Tensor(-0.052647278, shape=(), dtype=float32), tf.Tensor(0.0756154, shape=(), dtype=float32), tf.Tensor(0.020653268, shape=(), dtype=float32), tf.Tensor(-0.06249076, shape=(), dtype=float32), tf.Tensor(-0.036298044, shape=(), dtype=float32), tf.Tensor(0.011630775, shape=(), dtype=float32), tf.Tensor(0.068808146, shape=(), dtype=float32), tf.Tensor(-0.040713735, shape=(), dtype=float32), tf.Tensor(-0.01589599, shape=(), dtype=float32), tf.Tensor(0.022711592, shape=(), dtype=float32), tf.Tensor(0.010543281, shape=(), dtype=float32), tf.Tensor(-0.016693002, shape=(), dtype=float32), tf.Tensor(-0.0031882452, shape=(), dtype=float32), tf.Tensor(0.005573282, shape=(), dtype=float32), tf.Tensor(-0.038671143, shape=(), dtype=float32), tf.Tensor(-0.0023794014, shape=(), dtype=float32), tf.Tensor(-0.052908592, shape=(), dtype=float32), tf.Tensor(0.03668991, shape=(), dtype=float32), tf.Tensor(-0.013172233, shape=(), dtype=float32), tf.Tensor(-0.0239118, shape=(), dtype=float32), tf.Tensor(0.031588282, shape=(), dtype=float32), tf.Tensor(0.06316237, shape=(), dtype=float32), tf.Tensor(-0.017381046, shape=(), dtype=float32), tf.Tensor(0.04683652, shape=(), dtype=float32), tf.Tensor(0.027002355, shape=(), dtype=float32), tf.Tensor(0.041966658, shape=(), dtype=float32), tf.Tensor(0.095873885, shape=(), dtype=float32), tf.Tensor(0.06987687, shape=(), dtype=float32), tf.Tensor(-0.048182003, shape=(), dtype=float32), tf.Tensor(-0.02679722, shape=(), dtype=float32), tf.Tensor(0.01095338, shape=(), dtype=float32), tf.Tensor(0.0124590155, shape=(), dtype=float32), tf.Tensor(-0.03715376, shape=(), dtype=float32), tf.Tensor(0.011673558, shape=(), dtype=float32), tf.Tensor(0.025146147, shape=(), dtype=float32), tf.Tensor(0.005816766, shape=(), dtype=float32), tf.Tensor(-0.011063083, shape=(), dtype=float32), tf.Tensor(-0.0067656143, shape=(), dtype=float32), tf.Tensor(-0.07596833, shape=(), dtype=float32), tf.Tensor(0.07020325, shape=(), dtype=float32), tf.Tensor(0.07631031, shape=(), dtype=float32), tf.Tensor(0.050751124, shape=(), dtype=float32), tf.Tensor(-0.014694677, shape=(), dtype=float32), tf.Tensor(-0.009856238, shape=(), dtype=float32), tf.Tensor(0.0121922465, shape=(), dtype=float32), tf.Tensor(0.05429827, shape=(), dtype=float32), tf.Tensor(-0.02645367, shape=(), dtype=float32), tf.Tensor(0.013764614, shape=(), dtype=float32), tf.Tensor(-0.08916231, shape=(), dtype=float32), tf.Tensor(0.05923999, shape=(), dtype=float32), tf.Tensor(-0.042189505, shape=(), dtype=float32), tf.Tensor(-0.06420892, shape=(), dtype=float32), tf.Tensor(-0.01678818, shape=(), dtype=float32), tf.Tensor(0.015626308, shape=(), dtype=float32), tf.Tensor(-0.015270449, shape=(), dtype=float32), tf.Tensor(-0.030836768, shape=(), dtype=float32), tf.Tensor(-0.015567685, shape=(), dtype=float32), tf.Tensor(-0.0399256, shape=(), dtype=float32), tf.Tensor(0.05246063, shape=(), dtype=float32), tf.Tensor(-0.04651136, shape=(), dtype=float32), tf.Tensor(0.031117456, shape=(), dtype=float32), tf.Tensor(-0.023506302, shape=(), dtype=float32), tf.Tensor(-0.07396878, shape=(), dtype=float32), tf.Tensor(0.040847186, shape=(), dtype=float32), tf.Tensor(-0.0040281336, shape=(), dtype=float32), tf.Tensor(0.052818257, shape=(), dtype=float32), tf.Tensor(0.019324802, shape=(), dtype=float32), tf.Tensor(0.08227731, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300082</th>\n",
       "      <td>Averted. Diego and Captain Esteban once battled for the love of Florinda, and Esteban won. Florinda still wants Diego, but he's lost interest. Florinda   Why didn't I marry you instead of Esteban?  Don Diego   Probably because he asked you and I did not.</td>\n",
       "      <td>t13313</td>\n",
       "      <td>film</td>\n",
       "      <td>(tf.Tensor(0.016358314, shape=(), dtype=float32), tf.Tensor(-0.055671506, shape=(), dtype=float32), tf.Tensor(-0.029637381, shape=(), dtype=float32), tf.Tensor(-0.056957223, shape=(), dtype=float32), tf.Tensor(0.052265424, shape=(), dtype=float32), tf.Tensor(-0.008139698, shape=(), dtype=float32), tf.Tensor(-0.030265389, shape=(), dtype=float32), tf.Tensor(0.027607191, shape=(), dtype=float32), tf.Tensor(0.032367367, shape=(), dtype=float32), tf.Tensor(0.04106494, shape=(), dtype=float32), tf.Tensor(0.014060544, shape=(), dtype=float32), tf.Tensor(-0.0021166166, shape=(), dtype=float32), tf.Tensor(-0.08100254, shape=(), dtype=float32), tf.Tensor(0.030571435, shape=(), dtype=float32), tf.Tensor(0.035408, shape=(), dtype=float32), tf.Tensor(-0.03086979, shape=(), dtype=float32), tf.Tensor(0.04915643, shape=(), dtype=float32), tf.Tensor(-0.022131559, shape=(), dtype=float32), tf.Tensor(0.034481965, shape=(), dtype=float32), tf.Tensor(-0.0038810829, shape=(), dtype=float32), tf.Tensor(-0.028350623, shape=(), dtype=float32), tf.Tensor(-0.02234467, shape=(), dtype=float32), tf.Tensor(0.037224375, shape=(), dtype=float32), tf.Tensor(0.000751521, shape=(), dtype=float32), tf.Tensor(-0.042181917, shape=(), dtype=float32), tf.Tensor(-0.056799695, shape=(), dtype=float32), tf.Tensor(0.0901877, shape=(), dtype=float32), tf.Tensor(0.039094556, shape=(), dtype=float32), tf.Tensor(-0.011283439, shape=(), dtype=float32), tf.Tensor(-0.044154435, shape=(), dtype=float32), tf.Tensor(-0.04247438, shape=(), dtype=float32), tf.Tensor(0.0069788494, shape=(), dtype=float32), tf.Tensor(-0.030908998, shape=(), dtype=float32), tf.Tensor(-0.04946975, shape=(), dtype=float32), tf.Tensor(-0.046985492, shape=(), dtype=float32), tf.Tensor(-0.058152452, shape=(), dtype=float32), tf.Tensor(-0.039360806, shape=(), dtype=float32), tf.Tensor(-0.018504063, shape=(), dtype=float32), tf.Tensor(0.09645569, shape=(), dtype=float32), tf.Tensor(0.02572451, shape=(), dtype=float32), tf.Tensor(0.108570255, shape=(), dtype=float32), tf.Tensor(0.03272362, shape=(), dtype=float32), tf.Tensor(-0.020669617, shape=(), dtype=float32), tf.Tensor(0.03645965, shape=(), dtype=float32), tf.Tensor(0.006018922, shape=(), dtype=float32), tf.Tensor(0.034145925, shape=(), dtype=float32), tf.Tensor(-0.03645899, shape=(), dtype=float32), tf.Tensor(-0.06911441, shape=(), dtype=float32), tf.Tensor(0.0016972882, shape=(), dtype=float32), tf.Tensor(0.1351007, shape=(), dtype=float32), tf.Tensor(0.03229023, shape=(), dtype=float32), tf.Tensor(-0.039639413, shape=(), dtype=float32), tf.Tensor(0.029782634, shape=(), dtype=float32), tf.Tensor(0.04943118, shape=(), dtype=float32), tf.Tensor(0.014977691, shape=(), dtype=float32), tf.Tensor(0.047863957, shape=(), dtype=float32), tf.Tensor(-0.041000806, shape=(), dtype=float32), tf.Tensor(-0.041627176, shape=(), dtype=float32), tf.Tensor(0.02688666, shape=(), dtype=float32), tf.Tensor(0.04648368, shape=(), dtype=float32), tf.Tensor(0.04881043, shape=(), dtype=float32), tf.Tensor(-0.014024911, shape=(), dtype=float32), tf.Tensor(0.040062383, shape=(), dtype=float32), tf.Tensor(0.048952438, shape=(), dtype=float32), tf.Tensor(0.026217744, shape=(), dtype=float32), tf.Tensor(0.015916433, shape=(), dtype=float32), tf.Tensor(0.004321039, shape=(), dtype=float32), tf.Tensor(-0.020352382, shape=(), dtype=float32), tf.Tensor(0.029310487, shape=(), dtype=float32), tf.Tensor(-0.038597338, shape=(), dtype=float32), tf.Tensor(-0.036106978, shape=(), dtype=float32), tf.Tensor(-0.032617886, shape=(), dtype=float32), tf.Tensor(-0.00010633847, shape=(), dtype=float32), tf.Tensor(0.058943268, shape=(), dtype=float32), tf.Tensor(-0.031699147, shape=(), dtype=float32), tf.Tensor(0.024221651, shape=(), dtype=float32), tf.Tensor(-0.027113795, shape=(), dtype=float32), tf.Tensor(0.041020654, shape=(), dtype=float32), tf.Tensor(0.024591466, shape=(), dtype=float32), tf.Tensor(-0.062298473, shape=(), dtype=float32), tf.Tensor(0.01828891, shape=(), dtype=float32), tf.Tensor(0.044462077, shape=(), dtype=float32), tf.Tensor(0.010415857, shape=(), dtype=float32), tf.Tensor(-0.069450855, shape=(), dtype=float32), tf.Tensor(-0.008549134, shape=(), dtype=float32), tf.Tensor(0.020632012, shape=(), dtype=float32), tf.Tensor(-0.010715748, shape=(), dtype=float32), tf.Tensor(0.0095858, shape=(), dtype=float32), tf.Tensor(-0.05704763, shape=(), dtype=float32), tf.Tensor(-0.0030989805, shape=(), dtype=float32), tf.Tensor(-0.01156347, shape=(), dtype=float32), tf.Tensor(-0.018763319, shape=(), dtype=float32), tf.Tensor(0.04045568, shape=(), dtype=float32), tf.Tensor(0.0009379599, shape=(), dtype=float32), tf.Tensor(-0.07353291, shape=(), dtype=float32), tf.Tensor(0.08545972, shape=(), dtype=float32), tf.Tensor(0.09345397, shape=(), dtype=float32), tf.Tensor(-0.01154483, shape=(), dtype=float32), tf.Tensor(0.01290133, shape=(), dtype=float32), tf.Tensor(-0.022994723, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300083</th>\n",
       "      <td>And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!</td>\n",
       "      <td>t12565</td>\n",
       "      <td>film</td>\n",
       "      <td>(tf.Tensor(-0.07129504, shape=(), dtype=float32), tf.Tensor(0.05181775, shape=(), dtype=float32), tf.Tensor(-0.02489603, shape=(), dtype=float32), tf.Tensor(-0.023506913, shape=(), dtype=float32), tf.Tensor(0.01281394, shape=(), dtype=float32), tf.Tensor(0.08956977, shape=(), dtype=float32), tf.Tensor(0.037577093, shape=(), dtype=float32), tf.Tensor(0.051022176, shape=(), dtype=float32), tf.Tensor(-0.05336184, shape=(), dtype=float32), tf.Tensor(0.054678135, shape=(), dtype=float32), tf.Tensor(0.015588774, shape=(), dtype=float32), tf.Tensor(0.012392177, shape=(), dtype=float32), tf.Tensor(-0.008402479, shape=(), dtype=float32), tf.Tensor(0.023921112, shape=(), dtype=float32), tf.Tensor(0.01599294, shape=(), dtype=float32), tf.Tensor(0.012201126, shape=(), dtype=float32), tf.Tensor(0.12451981, shape=(), dtype=float32), tf.Tensor(-0.00078481203, shape=(), dtype=float32), tf.Tensor(0.021262959, shape=(), dtype=float32), tf.Tensor(-0.009026773, shape=(), dtype=float32), tf.Tensor(-0.005325233, shape=(), dtype=float32), tf.Tensor(-0.0049861637, shape=(), dtype=float32), tf.Tensor(0.070501365, shape=(), dtype=float32), tf.Tensor(-0.0037700974, shape=(), dtype=float32), tf.Tensor(0.034333933, shape=(), dtype=float32), tf.Tensor(0.02061151, shape=(), dtype=float32), tf.Tensor(-0.03962808, shape=(), dtype=float32), tf.Tensor(-0.013045015, shape=(), dtype=float32), tf.Tensor(0.02984808, shape=(), dtype=float32), tf.Tensor(-0.006684023, shape=(), dtype=float32), tf.Tensor(-0.10910741, shape=(), dtype=float32), tf.Tensor(0.049963016, shape=(), dtype=float32), tf.Tensor(0.010760641, shape=(), dtype=float32), tf.Tensor(0.0079353135, shape=(), dtype=float32), tf.Tensor(0.0815682, shape=(), dtype=float32), tf.Tensor(-0.014976661, shape=(), dtype=float32), tf.Tensor(0.018694662, shape=(), dtype=float32), tf.Tensor(-0.00011071347, shape=(), dtype=float32), tf.Tensor(0.015297066, shape=(), dtype=float32), tf.Tensor(0.017451882, shape=(), dtype=float32), tf.Tensor(-0.026891377, shape=(), dtype=float32), tf.Tensor(-0.060761075, shape=(), dtype=float32), tf.Tensor(-0.03561407, shape=(), dtype=float32), tf.Tensor(-0.060945064, shape=(), dtype=float32), tf.Tensor(0.08811272, shape=(), dtype=float32), tf.Tensor(0.03339591, shape=(), dtype=float32), tf.Tensor(0.046238407, shape=(), dtype=float32), tf.Tensor(-0.033360608, shape=(), dtype=float32), tf.Tensor(0.076256156, shape=(), dtype=float32), tf.Tensor(0.04862837, shape=(), dtype=float32), tf.Tensor(0.041696865, shape=(), dtype=float32), tf.Tensor(-0.051193308, shape=(), dtype=float32), tf.Tensor(-0.03433186, shape=(), dtype=float32), tf.Tensor(0.03920537, shape=(), dtype=float32), tf.Tensor(-0.05263466, shape=(), dtype=float32), tf.Tensor(-0.022765348, shape=(), dtype=float32), tf.Tensor(0.012549712, shape=(), dtype=float32), tf.Tensor(-0.042384088, shape=(), dtype=float32), tf.Tensor(-0.0037845653, shape=(), dtype=float32), tf.Tensor(0.038839225, shape=(), dtype=float32), tf.Tensor(-0.010468014, shape=(), dtype=float32), tf.Tensor(-0.013704077, shape=(), dtype=float32), tf.Tensor(-0.013660341, shape=(), dtype=float32), tf.Tensor(0.0622114, shape=(), dtype=float32), tf.Tensor(0.06779714, shape=(), dtype=float32), tf.Tensor(-0.016206617, shape=(), dtype=float32), tf.Tensor(-0.06212112, shape=(), dtype=float32), tf.Tensor(-0.021148777, shape=(), dtype=float32), tf.Tensor(-0.025203573, shape=(), dtype=float32), tf.Tensor(-0.0026878437, shape=(), dtype=float32), tf.Tensor(-0.038446885, shape=(), dtype=float32), tf.Tensor(-0.030847214, shape=(), dtype=float32), tf.Tensor(0.021817302, shape=(), dtype=float32), tf.Tensor(0.0075747157, shape=(), dtype=float32), tf.Tensor(0.016853029, shape=(), dtype=float32), tf.Tensor(0.01750912, shape=(), dtype=float32), tf.Tensor(0.024633944, shape=(), dtype=float32), tf.Tensor(-0.013615712, shape=(), dtype=float32), tf.Tensor(-0.0030897972, shape=(), dtype=float32), tf.Tensor(-0.008850621, shape=(), dtype=float32), tf.Tensor(-0.012036455, shape=(), dtype=float32), tf.Tensor(-0.022593455, shape=(), dtype=float32), tf.Tensor(-0.053547565, shape=(), dtype=float32), tf.Tensor(-0.080733314, shape=(), dtype=float32), tf.Tensor(0.0066507645, shape=(), dtype=float32), tf.Tensor(0.020736605, shape=(), dtype=float32), tf.Tensor(0.012811926, shape=(), dtype=float32), tf.Tensor(0.03420647, shape=(), dtype=float32), tf.Tensor(-0.030252958, shape=(), dtype=float32), tf.Tensor(-0.056493968, shape=(), dtype=float32), tf.Tensor(-0.017451148, shape=(), dtype=float32), tf.Tensor(-0.015234978, shape=(), dtype=float32), tf.Tensor(0.020664064, shape=(), dtype=float32), tf.Tensor(0.12766692, shape=(), dtype=float32), tf.Tensor(-0.07579299, shape=(), dtype=float32), tf.Tensor(-0.00022272844, shape=(), dtype=float32), tf.Tensor(0.028153209, shape=(), dtype=float32), tf.Tensor(-0.04204758, shape=(), dtype=float32), tf.Tensor(-0.10946826, shape=(), dtype=float32), tf.Tensor(-0.031043392, shape=(), dtype=float32), ...)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>288783 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                                                                                                                                                                                            Example  \\\n",
       "0       In A Bad Day For Voodoo,  Esmeralda, the maker of the voodoo dolls, explains to Tyler and Kelley that Adam is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "1                                                                When Tyler tries to   chase after his stolen car which has the voodoo doll of himself in it , he curses  \"S-word, f-word, s-word, d-word, s-word times three, f-word, and a z-word that [he] made up on the spot\".   \n",
       "2                           Esmeralda, the maker of the voodoo dolls, explains to  Tyler  and  Kelley  that  Adam  is this, and will one day defeat the hobgoblins. When Kelley questions this, she explains that only the chosen one could have made the voodoo dolls so powerful.   \n",
       "3                                                                                                                                                Tyler is the protagonist, and always tries to do the right thing, but states to the reader that he is definitely not a role model.   \n",
       "4                                                                                        Esmeralda the doll-maker tells  Tyler  and  Kelley ,  \"[ Adam ] add to power of doll. Make it super-magical. He  Chosen One .\", revealing that  Adam  of all people is  The Chosen One . .   \n",
       "...                                                                                                                                                                                                                                                                             ...   \n",
       "300079                                                                                                                                                   All the caballeros are California placenames with \"Don\" instead of \"San\"  Diego, Francisco, Fernando, Jose,  Luis Obispo .   \n",
       "300080                                                                                                                                                                                                                                                                   \"Sink me!\"   \n",
       "300081                                                                                                                                        Diego jumps from a second-floor window to escape pursuit ... and injures himself severely enough to force his Zorroing into a siesta.   \n",
       "300082               Averted. Diego and Captain Esteban once battled for the love of Florinda, and Esteban won. Florinda still wants Diego, but he's lost interest. Florinda   Why didn't I marry you instead of Esteban?  Don Diego   Probably because he asked you and I did not.   \n",
       "300083                                                And how! Between Hamilton's over the top campiness, Vaccaro's ditziness and Liebman's outrageous overacting, this movie is practically a buffet of ham. Even Donovon Scott comes across like this, and his character is mute!   \n",
       "\n",
       "       trope_id example_type  \\\n",
       "0        t22727          lit   \n",
       "1        t01294          lit   \n",
       "2        t22727          lit   \n",
       "3        t01080          lit   \n",
       "4        t25818          lit   \n",
       "...         ...          ...   \n",
       "300079   t23159         film   \n",
       "300080   t03382         film   \n",
       "300081   t18297         film   \n",
       "300082   t13313         film   \n",
       "300083   t12565         film   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  embedding  \n",
       "0            (tf.Tensor(0.08800587, shape=(), dtype=float32), tf.Tensor(-0.08599387, shape=(), dtype=float32), tf.Tensor(-0.025380103, shape=(), dtype=float32), tf.Tensor(0.051938705, shape=(), dtype=float32), tf.Tensor(0.02031321, shape=(), dtype=float32), tf.Tensor(-0.0020598548, shape=(), dtype=float32), tf.Tensor(0.01301749, shape=(), dtype=float32), tf.Tensor(0.04751223, shape=(), dtype=float32), tf.Tensor(-0.0019549562, shape=(), dtype=float32), tf.Tensor(-0.05235701, shape=(), dtype=float32), tf.Tensor(-0.028665619, shape=(), dtype=float32), tf.Tensor(0.040233985, shape=(), dtype=float32), tf.Tensor(-0.0012896181, shape=(), dtype=float32), tf.Tensor(0.043469764, shape=(), dtype=float32), tf.Tensor(0.04371827, shape=(), dtype=float32), tf.Tensor(-0.012290332, shape=(), dtype=float32), tf.Tensor(0.08683354, shape=(), dtype=float32), tf.Tensor(0.021370953, shape=(), dtype=float32), tf.Tensor(-0.00017330215, shape=(), dtype=float32), tf.Tensor(-0.036185037, shape=(), dtype=float32), tf.Tensor(-0.04546309, shape=(), dtype=float32), tf.Tensor(0.037026417, shape=(), dtype=float32), tf.Tensor(0.058535624, shape=(), dtype=float32), tf.Tensor(-0.062196318, shape=(), dtype=float32), tf.Tensor(0.004359815, shape=(), dtype=float32), tf.Tensor(-0.002621128, shape=(), dtype=float32), tf.Tensor(-0.0023814337, shape=(), dtype=float32), tf.Tensor(0.079657964, shape=(), dtype=float32), tf.Tensor(0.016618242, shape=(), dtype=float32), tf.Tensor(-0.057619054, shape=(), dtype=float32), tf.Tensor(-0.06499728, shape=(), dtype=float32), tf.Tensor(-0.032943424, shape=(), dtype=float32), tf.Tensor(0.04166872, shape=(), dtype=float32), tf.Tensor(-0.0692408, shape=(), dtype=float32), tf.Tensor(0.004227797, shape=(), dtype=float32), tf.Tensor(-0.022902718, shape=(), dtype=float32), tf.Tensor(0.02059966, shape=(), dtype=float32), tf.Tensor(-0.00035657547, shape=(), dtype=float32), tf.Tensor(-0.029187722, shape=(), dtype=float32), tf.Tensor(-0.022358175, shape=(), dtype=float32), tf.Tensor(-0.022271976, shape=(), dtype=float32), tf.Tensor(0.04697008, shape=(), dtype=float32), tf.Tensor(-0.07002722, shape=(), dtype=float32), tf.Tensor(-0.027185066, shape=(), dtype=float32), tf.Tensor(-0.0030682483, shape=(), dtype=float32), tf.Tensor(-0.030895032, shape=(), dtype=float32), tf.Tensor(0.079647385, shape=(), dtype=float32), tf.Tensor(0.057742752, shape=(), dtype=float32), tf.Tensor(-0.025763107, shape=(), dtype=float32), tf.Tensor(0.07856635, shape=(), dtype=float32), tf.Tensor(-0.021441985, shape=(), dtype=float32), tf.Tensor(-0.0592378, shape=(), dtype=float32), tf.Tensor(0.00816241, shape=(), dtype=float32), tf.Tensor(-0.046447586, shape=(), dtype=float32), tf.Tensor(0.02299255, shape=(), dtype=float32), tf.Tensor(0.044731148, shape=(), dtype=float32), tf.Tensor(-0.053904887, shape=(), dtype=float32), tf.Tensor(0.008835485, shape=(), dtype=float32), tf.Tensor(0.031931505, shape=(), dtype=float32), tf.Tensor(0.00012375598, shape=(), dtype=float32), tf.Tensor(-0.009230669, shape=(), dtype=float32), tf.Tensor(-0.031224066, shape=(), dtype=float32), tf.Tensor(-0.02890986, shape=(), dtype=float32), tf.Tensor(0.053919934, shape=(), dtype=float32), tf.Tensor(-0.008776098, shape=(), dtype=float32), tf.Tensor(-0.030362194, shape=(), dtype=float32), tf.Tensor(-0.055678792, shape=(), dtype=float32), tf.Tensor(-0.017249918, shape=(), dtype=float32), tf.Tensor(-0.006067661, shape=(), dtype=float32), tf.Tensor(0.019166904, shape=(), dtype=float32), tf.Tensor(-0.023110569, shape=(), dtype=float32), tf.Tensor(-0.055081792, shape=(), dtype=float32), tf.Tensor(-0.0037445223, shape=(), dtype=float32), tf.Tensor(0.009316371, shape=(), dtype=float32), tf.Tensor(-0.06440608, shape=(), dtype=float32), tf.Tensor(-0.010236327, shape=(), dtype=float32), tf.Tensor(0.060167458, shape=(), dtype=float32), tf.Tensor(0.045644686, shape=(), dtype=float32), tf.Tensor(0.08203374, shape=(), dtype=float32), tf.Tensor(0.020977907, shape=(), dtype=float32), tf.Tensor(-0.07048969, shape=(), dtype=float32), tf.Tensor(-0.03973232, shape=(), dtype=float32), tf.Tensor(-0.009789303, shape=(), dtype=float32), tf.Tensor(-0.05202683, shape=(), dtype=float32), tf.Tensor(-0.054066382, shape=(), dtype=float32), tf.Tensor(-0.02701431, shape=(), dtype=float32), tf.Tensor(0.009901098, shape=(), dtype=float32), tf.Tensor(-0.02209809, shape=(), dtype=float32), tf.Tensor(0.08363872, shape=(), dtype=float32), tf.Tensor(0.008711273, shape=(), dtype=float32), tf.Tensor(-0.056222685, shape=(), dtype=float32), tf.Tensor(-0.022011213, shape=(), dtype=float32), tf.Tensor(0.053444725, shape=(), dtype=float32), tf.Tensor(0.01691654, shape=(), dtype=float32), tf.Tensor(-0.07186151, shape=(), dtype=float32), tf.Tensor(-0.05782236, shape=(), dtype=float32), tf.Tensor(0.007892808, shape=(), dtype=float32), tf.Tensor(-0.0030549834, shape=(), dtype=float32), tf.Tensor(-0.04067482, shape=(), dtype=float32), tf.Tensor(0.011997106, shape=(), dtype=float32), ...)  \n",
       "1                    (tf.Tensor(0.010330548, shape=(), dtype=float32), tf.Tensor(-0.029361794, shape=(), dtype=float32), tf.Tensor(-0.02274592, shape=(), dtype=float32), tf.Tensor(-0.002653622, shape=(), dtype=float32), tf.Tensor(-0.030094229, shape=(), dtype=float32), tf.Tensor(0.07554659, shape=(), dtype=float32), tf.Tensor(0.0106902355, shape=(), dtype=float32), tf.Tensor(-0.010981429, shape=(), dtype=float32), tf.Tensor(0.026712697, shape=(), dtype=float32), tf.Tensor(0.026856365, shape=(), dtype=float32), tf.Tensor(-0.025904251, shape=(), dtype=float32), tf.Tensor(0.0047908076, shape=(), dtype=float32), tf.Tensor(0.0050791674, shape=(), dtype=float32), tf.Tensor(0.04752505, shape=(), dtype=float32), tf.Tensor(0.012012608, shape=(), dtype=float32), tf.Tensor(-0.004147278, shape=(), dtype=float32), tf.Tensor(0.130424, shape=(), dtype=float32), tf.Tensor(-0.01596583, shape=(), dtype=float32), tf.Tensor(0.007897065, shape=(), dtype=float32), tf.Tensor(-0.028737511, shape=(), dtype=float32), tf.Tensor(-0.060379915, shape=(), dtype=float32), tf.Tensor(0.047210198, shape=(), dtype=float32), tf.Tensor(0.062640674, shape=(), dtype=float32), tf.Tensor(0.019653106, shape=(), dtype=float32), tf.Tensor(-0.05946766, shape=(), dtype=float32), tf.Tensor(0.061379302, shape=(), dtype=float32), tf.Tensor(-0.08129871, shape=(), dtype=float32), tf.Tensor(0.0068943007, shape=(), dtype=float32), tf.Tensor(0.03735557, shape=(), dtype=float32), tf.Tensor(-0.04145611, shape=(), dtype=float32), tf.Tensor(-0.124526694, shape=(), dtype=float32), tf.Tensor(-0.0046156486, shape=(), dtype=float32), tf.Tensor(-0.019345704, shape=(), dtype=float32), tf.Tensor(0.04755795, shape=(), dtype=float32), tf.Tensor(-0.04352252, shape=(), dtype=float32), tf.Tensor(-0.023820508, shape=(), dtype=float32), tf.Tensor(-0.01023172, shape=(), dtype=float32), tf.Tensor(0.0007520435, shape=(), dtype=float32), tf.Tensor(-0.0076776952, shape=(), dtype=float32), tf.Tensor(-0.057130553, shape=(), dtype=float32), tf.Tensor(-0.0026168914, shape=(), dtype=float32), tf.Tensor(-0.023852542, shape=(), dtype=float32), tf.Tensor(-0.026527165, shape=(), dtype=float32), tf.Tensor(0.00721357, shape=(), dtype=float32), tf.Tensor(-0.048666157, shape=(), dtype=float32), tf.Tensor(0.017667599, shape=(), dtype=float32), tf.Tensor(0.05699326, shape=(), dtype=float32), tf.Tensor(0.010426503, shape=(), dtype=float32), tf.Tensor(0.01827146, shape=(), dtype=float32), tf.Tensor(0.02974568, shape=(), dtype=float32), tf.Tensor(-0.0019398523, shape=(), dtype=float32), tf.Tensor(-0.015875578, shape=(), dtype=float32), tf.Tensor(0.015058477, shape=(), dtype=float32), tf.Tensor(0.013088203, shape=(), dtype=float32), tf.Tensor(0.04991926, shape=(), dtype=float32), tf.Tensor(0.043303464, shape=(), dtype=float32), tf.Tensor(-0.015027079, shape=(), dtype=float32), tf.Tensor(-0.020142818, shape=(), dtype=float32), tf.Tensor(-0.01281049, shape=(), dtype=float32), tf.Tensor(0.07224833, shape=(), dtype=float32), tf.Tensor(-0.024062173, shape=(), dtype=float32), tf.Tensor(-0.054629564, shape=(), dtype=float32), tf.Tensor(-0.076675266, shape=(), dtype=float32), tf.Tensor(-0.031861987, shape=(), dtype=float32), tf.Tensor(0.035590075, shape=(), dtype=float32), tf.Tensor(0.043255467, shape=(), dtype=float32), tf.Tensor(-0.060752455, shape=(), dtype=float32), tf.Tensor(-0.03541548, shape=(), dtype=float32), tf.Tensor(0.03134576, shape=(), dtype=float32), tf.Tensor(0.0007455198, shape=(), dtype=float32), tf.Tensor(0.0042349533, shape=(), dtype=float32), tf.Tensor(0.010123341, shape=(), dtype=float32), tf.Tensor(0.059565395, shape=(), dtype=float32), tf.Tensor(-0.009693012, shape=(), dtype=float32), tf.Tensor(-0.017010922, shape=(), dtype=float32), tf.Tensor(-0.015895218, shape=(), dtype=float32), tf.Tensor(0.12877099, shape=(), dtype=float32), tf.Tensor(0.06468694, shape=(), dtype=float32), tf.Tensor(0.018707115, shape=(), dtype=float32), tf.Tensor(0.009936, shape=(), dtype=float32), tf.Tensor(-0.0043845777, shape=(), dtype=float32), tf.Tensor(-0.024331015, shape=(), dtype=float32), tf.Tensor(-0.036014624, shape=(), dtype=float32), tf.Tensor(-0.06364599, shape=(), dtype=float32), tf.Tensor(-0.048503984, shape=(), dtype=float32), tf.Tensor(-0.050774828, shape=(), dtype=float32), tf.Tensor(0.029971149, shape=(), dtype=float32), tf.Tensor(-0.024143403, shape=(), dtype=float32), tf.Tensor(0.06282789, shape=(), dtype=float32), tf.Tensor(-0.028649233, shape=(), dtype=float32), tf.Tensor(0.016530229, shape=(), dtype=float32), tf.Tensor(-0.019938834, shape=(), dtype=float32), tf.Tensor(0.06232069, shape=(), dtype=float32), tf.Tensor(0.045510154, shape=(), dtype=float32), tf.Tensor(-0.05315823, shape=(), dtype=float32), tf.Tensor(-0.09995444, shape=(), dtype=float32), tf.Tensor(-0.05620101, shape=(), dtype=float32), tf.Tensor(-0.01500156, shape=(), dtype=float32), tf.Tensor(-0.01709219, shape=(), dtype=float32), tf.Tensor(0.027486103, shape=(), dtype=float32), ...)  \n",
       "2              (tf.Tensor(0.07903765, shape=(), dtype=float32), tf.Tensor(-0.08122551, shape=(), dtype=float32), tf.Tensor(-0.029343614, shape=(), dtype=float32), tf.Tensor(0.03724364, shape=(), dtype=float32), tf.Tensor(0.011438731, shape=(), dtype=float32), tf.Tensor(-0.002793295, shape=(), dtype=float32), tf.Tensor(0.017981712, shape=(), dtype=float32), tf.Tensor(0.06524445, shape=(), dtype=float32), tf.Tensor(-0.018213881, shape=(), dtype=float32), tf.Tensor(-0.06824911, shape=(), dtype=float32), tf.Tensor(-0.03127585, shape=(), dtype=float32), tf.Tensor(0.06952839, shape=(), dtype=float32), tf.Tensor(-0.017614137, shape=(), dtype=float32), tf.Tensor(0.055293288, shape=(), dtype=float32), tf.Tensor(0.040941577, shape=(), dtype=float32), tf.Tensor(-0.023553396, shape=(), dtype=float32), tf.Tensor(0.09257069, shape=(), dtype=float32), tf.Tensor(0.022809014, shape=(), dtype=float32), tf.Tensor(0.007699889, shape=(), dtype=float32), tf.Tensor(-0.012788549, shape=(), dtype=float32), tf.Tensor(-0.04757676, shape=(), dtype=float32), tf.Tensor(0.028746195, shape=(), dtype=float32), tf.Tensor(0.05641736, shape=(), dtype=float32), tf.Tensor(-0.054203033, shape=(), dtype=float32), tf.Tensor(-0.010500583, shape=(), dtype=float32), tf.Tensor(-0.004929476, shape=(), dtype=float32), tf.Tensor(-0.006849469, shape=(), dtype=float32), tf.Tensor(0.061456073, shape=(), dtype=float32), tf.Tensor(0.023601279, shape=(), dtype=float32), tf.Tensor(-0.05851914, shape=(), dtype=float32), tf.Tensor(-0.063357376, shape=(), dtype=float32), tf.Tensor(-0.050873302, shape=(), dtype=float32), tf.Tensor(0.053865567, shape=(), dtype=float32), tf.Tensor(-0.059307277, shape=(), dtype=float32), tf.Tensor(0.0042104274, shape=(), dtype=float32), tf.Tensor(-0.035044774, shape=(), dtype=float32), tf.Tensor(0.013260701, shape=(), dtype=float32), tf.Tensor(-0.0073710675, shape=(), dtype=float32), tf.Tensor(-0.022457961, shape=(), dtype=float32), tf.Tensor(-0.02190834, shape=(), dtype=float32), tf.Tensor(-0.02646201, shape=(), dtype=float32), tf.Tensor(0.05482588, shape=(), dtype=float32), tf.Tensor(-0.050045248, shape=(), dtype=float32), tf.Tensor(-0.026658569, shape=(), dtype=float32), tf.Tensor(-0.0026801967, shape=(), dtype=float32), tf.Tensor(-0.022143172, shape=(), dtype=float32), tf.Tensor(0.08452685, shape=(), dtype=float32), tf.Tensor(0.052563205, shape=(), dtype=float32), tf.Tensor(-0.028089037, shape=(), dtype=float32), tf.Tensor(0.048151925, shape=(), dtype=float32), tf.Tensor(-0.01412995, shape=(), dtype=float32), tf.Tensor(-0.04600642, shape=(), dtype=float32), tf.Tensor(0.014064551, shape=(), dtype=float32), tf.Tensor(-0.06108639, shape=(), dtype=float32), tf.Tensor(0.014461201, shape=(), dtype=float32), tf.Tensor(0.039829098, shape=(), dtype=float32), tf.Tensor(-0.038568724, shape=(), dtype=float32), tf.Tensor(0.003268224, shape=(), dtype=float32), tf.Tensor(0.029916162, shape=(), dtype=float32), tf.Tensor(-0.011676105, shape=(), dtype=float32), tf.Tensor(-0.013288738, shape=(), dtype=float32), tf.Tensor(-0.037835382, shape=(), dtype=float32), tf.Tensor(-0.048727572, shape=(), dtype=float32), tf.Tensor(0.063617, shape=(), dtype=float32), tf.Tensor(-0.018076634, shape=(), dtype=float32), tf.Tensor(-0.01990476, shape=(), dtype=float32), tf.Tensor(-0.042753935, shape=(), dtype=float32), tf.Tensor(-0.018533094, shape=(), dtype=float32), tf.Tensor(-0.00074974366, shape=(), dtype=float32), tf.Tensor(0.0008349011, shape=(), dtype=float32), tf.Tensor(-0.012258501, shape=(), dtype=float32), tf.Tensor(-0.02587158, shape=(), dtype=float32), tf.Tensor(-0.0052914466, shape=(), dtype=float32), tf.Tensor(0.00012988648, shape=(), dtype=float32), tf.Tensor(-0.051806636, shape=(), dtype=float32), tf.Tensor(-0.0068446877, shape=(), dtype=float32), tf.Tensor(0.045981146, shape=(), dtype=float32), tf.Tensor(0.014970127, shape=(), dtype=float32), tf.Tensor(0.07396664, shape=(), dtype=float32), tf.Tensor(0.0449271, shape=(), dtype=float32), tf.Tensor(-0.05091219, shape=(), dtype=float32), tf.Tensor(-0.03671727, shape=(), dtype=float32), tf.Tensor(-0.027304694, shape=(), dtype=float32), tf.Tensor(-0.04140609, shape=(), dtype=float32), tf.Tensor(-0.0636552, shape=(), dtype=float32), tf.Tensor(-0.032051597, shape=(), dtype=float32), tf.Tensor(0.023820529, shape=(), dtype=float32), tf.Tensor(-0.0074291355, shape=(), dtype=float32), tf.Tensor(0.09682213, shape=(), dtype=float32), tf.Tensor(0.020731114, shape=(), dtype=float32), tf.Tensor(-0.07162586, shape=(), dtype=float32), tf.Tensor(-0.017131131, shape=(), dtype=float32), tf.Tensor(0.05679401, shape=(), dtype=float32), tf.Tensor(0.020841591, shape=(), dtype=float32), tf.Tensor(-0.07075116, shape=(), dtype=float32), tf.Tensor(-0.06193501, shape=(), dtype=float32), tf.Tensor(0.006631097, shape=(), dtype=float32), tf.Tensor(-0.0039047645, shape=(), dtype=float32), tf.Tensor(-0.04713479, shape=(), dtype=float32), tf.Tensor(-0.015411815, shape=(), dtype=float32), ...)  \n",
       "3                                  (tf.Tensor(-0.03549712, shape=(), dtype=float32), tf.Tensor(0.03641831, shape=(), dtype=float32), tf.Tensor(0.018138606, shape=(), dtype=float32), tf.Tensor(0.04770905, shape=(), dtype=float32), tf.Tensor(-0.004124664, shape=(), dtype=float32), tf.Tensor(0.0024107983, shape=(), dtype=float32), tf.Tensor(-0.040756386, shape=(), dtype=float32), tf.Tensor(-0.034914747, shape=(), dtype=float32), tf.Tensor(0.020394739, shape=(), dtype=float32), tf.Tensor(0.0172372, shape=(), dtype=float32), tf.Tensor(-0.041197516, shape=(), dtype=float32), tf.Tensor(-0.026053572, shape=(), dtype=float32), tf.Tensor(-0.02246861, shape=(), dtype=float32), tf.Tensor(0.017176801, shape=(), dtype=float32), tf.Tensor(0.032556154, shape=(), dtype=float32), tf.Tensor(0.021616759, shape=(), dtype=float32), tf.Tensor(0.136648, shape=(), dtype=float32), tf.Tensor(0.05070697, shape=(), dtype=float32), tf.Tensor(-0.0370852, shape=(), dtype=float32), tf.Tensor(0.028701453, shape=(), dtype=float32), tf.Tensor(-0.059691045, shape=(), dtype=float32), tf.Tensor(0.010044211, shape=(), dtype=float32), tf.Tensor(0.059767265, shape=(), dtype=float32), tf.Tensor(0.011906863, shape=(), dtype=float32), tf.Tensor(-0.00084792287, shape=(), dtype=float32), tf.Tensor(0.01737797, shape=(), dtype=float32), tf.Tensor(-0.011986313, shape=(), dtype=float32), tf.Tensor(0.10150425, shape=(), dtype=float32), tf.Tensor(0.07557448, shape=(), dtype=float32), tf.Tensor(-0.037511464, shape=(), dtype=float32), tf.Tensor(-0.0392109, shape=(), dtype=float32), tf.Tensor(-0.026490528, shape=(), dtype=float32), tf.Tensor(-0.05747621, shape=(), dtype=float32), tf.Tensor(0.058619354, shape=(), dtype=float32), tf.Tensor(-0.06631421, shape=(), dtype=float32), tf.Tensor(-0.051940884, shape=(), dtype=float32), tf.Tensor(0.02359717, shape=(), dtype=float32), tf.Tensor(0.023845933, shape=(), dtype=float32), tf.Tensor(0.055003658, shape=(), dtype=float32), tf.Tensor(-0.0891166, shape=(), dtype=float32), tf.Tensor(0.047206674, shape=(), dtype=float32), tf.Tensor(-0.04001871, shape=(), dtype=float32), tf.Tensor(0.046865817, shape=(), dtype=float32), tf.Tensor(0.003833754, shape=(), dtype=float32), tf.Tensor(-0.050639827, shape=(), dtype=float32), tf.Tensor(-0.009351717, shape=(), dtype=float32), tf.Tensor(0.06790367, shape=(), dtype=float32), tf.Tensor(0.028699124, shape=(), dtype=float32), tf.Tensor(0.036173567, shape=(), dtype=float32), tf.Tensor(-0.017418522, shape=(), dtype=float32), tf.Tensor(0.00669771, shape=(), dtype=float32), tf.Tensor(-0.05520239, shape=(), dtype=float32), tf.Tensor(-0.037526235, shape=(), dtype=float32), tf.Tensor(0.0018540716, shape=(), dtype=float32), tf.Tensor(-0.032877833, shape=(), dtype=float32), tf.Tensor(-0.0038160027, shape=(), dtype=float32), tf.Tensor(0.04226888, shape=(), dtype=float32), tf.Tensor(-0.010834188, shape=(), dtype=float32), tf.Tensor(0.06358276, shape=(), dtype=float32), tf.Tensor(-0.016166119, shape=(), dtype=float32), tf.Tensor(-0.013715793, shape=(), dtype=float32), tf.Tensor(-0.024011398, shape=(), dtype=float32), tf.Tensor(-0.0028317412, shape=(), dtype=float32), tf.Tensor(-0.012698178, shape=(), dtype=float32), tf.Tensor(-0.06302096, shape=(), dtype=float32), tf.Tensor(0.0494479, shape=(), dtype=float32), tf.Tensor(-0.01340376, shape=(), dtype=float32), tf.Tensor(-0.023631789, shape=(), dtype=float32), tf.Tensor(0.09890174, shape=(), dtype=float32), tf.Tensor(-0.08014861, shape=(), dtype=float32), tf.Tensor(0.06290152, shape=(), dtype=float32), tf.Tensor(0.031284846, shape=(), dtype=float32), tf.Tensor(-0.06275099, shape=(), dtype=float32), tf.Tensor(0.041656412, shape=(), dtype=float32), tf.Tensor(0.036877573, shape=(), dtype=float32), tf.Tensor(0.033646654, shape=(), dtype=float32), tf.Tensor(0.039049953, shape=(), dtype=float32), tf.Tensor(-0.059652343, shape=(), dtype=float32), tf.Tensor(0.03144189, shape=(), dtype=float32), tf.Tensor(-0.016832195, shape=(), dtype=float32), tf.Tensor(-0.05423973, shape=(), dtype=float32), tf.Tensor(-0.046023183, shape=(), dtype=float32), tf.Tensor(-0.075923145, shape=(), dtype=float32), tf.Tensor(-0.03515708, shape=(), dtype=float32), tf.Tensor(-0.020270295, shape=(), dtype=float32), tf.Tensor(0.0070862304, shape=(), dtype=float32), tf.Tensor(0.014913817, shape=(), dtype=float32), tf.Tensor(0.059813924, shape=(), dtype=float32), tf.Tensor(0.020985503, shape=(), dtype=float32), tf.Tensor(-0.026704855, shape=(), dtype=float32), tf.Tensor(-0.09044281, shape=(), dtype=float32), tf.Tensor(0.0038491231, shape=(), dtype=float32), tf.Tensor(0.046002652, shape=(), dtype=float32), tf.Tensor(0.07362481, shape=(), dtype=float32), tf.Tensor(-0.058121517, shape=(), dtype=float32), tf.Tensor(-0.034743812, shape=(), dtype=float32), tf.Tensor(0.03586781, shape=(), dtype=float32), tf.Tensor(-0.02638692, shape=(), dtype=float32), tf.Tensor(-0.07126451, shape=(), dtype=float32), tf.Tensor(-0.012174028, shape=(), dtype=float32), ...)  \n",
       "4       (tf.Tensor(0.023967337, shape=(), dtype=float32), tf.Tensor(-0.084397495, shape=(), dtype=float32), tf.Tensor(0.01415953, shape=(), dtype=float32), tf.Tensor(-0.01036743, shape=(), dtype=float32), tf.Tensor(0.035001464, shape=(), dtype=float32), tf.Tensor(0.005071171, shape=(), dtype=float32), tf.Tensor(-0.01694777, shape=(), dtype=float32), tf.Tensor(-0.03718967, shape=(), dtype=float32), tf.Tensor(0.024154918, shape=(), dtype=float32), tf.Tensor(-0.050603937, shape=(), dtype=float32), tf.Tensor(-0.005885336, shape=(), dtype=float32), tf.Tensor(0.07657696, shape=(), dtype=float32), tf.Tensor(-0.05108257, shape=(), dtype=float32), tf.Tensor(0.03954993, shape=(), dtype=float32), tf.Tensor(0.057510085, shape=(), dtype=float32), tf.Tensor(-0.018536614, shape=(), dtype=float32), tf.Tensor(0.12386522, shape=(), dtype=float32), tf.Tensor(0.032914054, shape=(), dtype=float32), tf.Tensor(-0.006091986, shape=(), dtype=float32), tf.Tensor(-0.0071318448, shape=(), dtype=float32), tf.Tensor(-0.07268326, shape=(), dtype=float32), tf.Tensor(-0.01380359, shape=(), dtype=float32), tf.Tensor(0.018614434, shape=(), dtype=float32), tf.Tensor(-0.0332584, shape=(), dtype=float32), tf.Tensor(-0.026341196, shape=(), dtype=float32), tf.Tensor(-0.021997431, shape=(), dtype=float32), tf.Tensor(0.0015622203, shape=(), dtype=float32), tf.Tensor(0.044560604, shape=(), dtype=float32), tf.Tensor(-0.015563141, shape=(), dtype=float32), tf.Tensor(-0.05528924, shape=(), dtype=float32), tf.Tensor(-0.069653936, shape=(), dtype=float32), tf.Tensor(-0.037985787, shape=(), dtype=float32), tf.Tensor(0.066987574, shape=(), dtype=float32), tf.Tensor(0.0042458903, shape=(), dtype=float32), tf.Tensor(-0.0039066095, shape=(), dtype=float32), tf.Tensor(0.00080658443, shape=(), dtype=float32), tf.Tensor(-0.018237745, shape=(), dtype=float32), tf.Tensor(0.013343024, shape=(), dtype=float32), tf.Tensor(-0.031679623, shape=(), dtype=float32), tf.Tensor(-0.021655528, shape=(), dtype=float32), tf.Tensor(0.018567353, shape=(), dtype=float32), tf.Tensor(0.055041708, shape=(), dtype=float32), tf.Tensor(-0.064987734, shape=(), dtype=float32), tf.Tensor(-0.0026901325, shape=(), dtype=float32), tf.Tensor(-0.029532868, shape=(), dtype=float32), tf.Tensor(-0.043519426, shape=(), dtype=float32), tf.Tensor(0.048361097, shape=(), dtype=float32), tf.Tensor(0.026961464, shape=(), dtype=float32), tf.Tensor(0.022430457, shape=(), dtype=float32), tf.Tensor(0.054570664, shape=(), dtype=float32), tf.Tensor(0.0007716413, shape=(), dtype=float32), tf.Tensor(-0.023416478, shape=(), dtype=float32), tf.Tensor(0.036561195, shape=(), dtype=float32), tf.Tensor(-0.05937416, shape=(), dtype=float32), tf.Tensor(-0.045613628, shape=(), dtype=float32), tf.Tensor(0.060605895, shape=(), dtype=float32), tf.Tensor(0.025672719, shape=(), dtype=float32), tf.Tensor(-0.00018210869, shape=(), dtype=float32), tf.Tensor(0.04163569, shape=(), dtype=float32), tf.Tensor(0.029721607, shape=(), dtype=float32), tf.Tensor(0.005982857, shape=(), dtype=float32), tf.Tensor(-0.033309005, shape=(), dtype=float32), tf.Tensor(-0.05784923, shape=(), dtype=float32), tf.Tensor(0.039723407, shape=(), dtype=float32), tf.Tensor(-0.04391734, shape=(), dtype=float32), tf.Tensor(0.0036102408, shape=(), dtype=float32), tf.Tensor(-0.052760348, shape=(), dtype=float32), tf.Tensor(-0.021783441, shape=(), dtype=float32), tf.Tensor(-0.004423645, shape=(), dtype=float32), tf.Tensor(-0.011782224, shape=(), dtype=float32), tf.Tensor(0.0259689, shape=(), dtype=float32), tf.Tensor(-0.020055423, shape=(), dtype=float32), tf.Tensor(-0.01234067, shape=(), dtype=float32), tf.Tensor(0.020949947, shape=(), dtype=float32), tf.Tensor(-0.029320963, shape=(), dtype=float32), tf.Tensor(0.009272562, shape=(), dtype=float32), tf.Tensor(-0.004329086, shape=(), dtype=float32), tf.Tensor(-0.017080763, shape=(), dtype=float32), tf.Tensor(0.057862487, shape=(), dtype=float32), tf.Tensor(0.04540068, shape=(), dtype=float32), tf.Tensor(-0.048767716, shape=(), dtype=float32), tf.Tensor(-0.014069336, shape=(), dtype=float32), tf.Tensor(-0.018453393, shape=(), dtype=float32), tf.Tensor(-0.045883752, shape=(), dtype=float32), tf.Tensor(-0.053496797, shape=(), dtype=float32), tf.Tensor(0.03852861, shape=(), dtype=float32), tf.Tensor(0.03908949, shape=(), dtype=float32), tf.Tensor(-0.013852586, shape=(), dtype=float32), tf.Tensor(0.03313901, shape=(), dtype=float32), tf.Tensor(0.026762994, shape=(), dtype=float32), tf.Tensor(-0.103191234, shape=(), dtype=float32), tf.Tensor(-0.021607641, shape=(), dtype=float32), tf.Tensor(0.035765253, shape=(), dtype=float32), tf.Tensor(0.03716235, shape=(), dtype=float32), tf.Tensor(-0.040432636, shape=(), dtype=float32), tf.Tensor(0.0053194156, shape=(), dtype=float32), tf.Tensor(0.020130217, shape=(), dtype=float32), tf.Tensor(-0.011448691, shape=(), dtype=float32), tf.Tensor(-0.030330446, shape=(), dtype=float32), tf.Tensor(-0.039985277, shape=(), dtype=float32), ...)  \n",
       "...                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     ...  \n",
       "300079     (tf.Tensor(-0.056905918, shape=(), dtype=float32), tf.Tensor(-0.029011853, shape=(), dtype=float32), tf.Tensor(-0.032758486, shape=(), dtype=float32), tf.Tensor(-0.033120498, shape=(), dtype=float32), tf.Tensor(-0.016638504, shape=(), dtype=float32), tf.Tensor(-0.03368108, shape=(), dtype=float32), tf.Tensor(0.028499708, shape=(), dtype=float32), tf.Tensor(-0.01722193, shape=(), dtype=float32), tf.Tensor(-0.059155736, shape=(), dtype=float32), tf.Tensor(0.06507286, shape=(), dtype=float32), tf.Tensor(0.02724566, shape=(), dtype=float32), tf.Tensor(-0.0013793716, shape=(), dtype=float32), tf.Tensor(-0.103186354, shape=(), dtype=float32), tf.Tensor(0.083762184, shape=(), dtype=float32), tf.Tensor(0.04576251, shape=(), dtype=float32), tf.Tensor(-0.058948424, shape=(), dtype=float32), tf.Tensor(0.074407704, shape=(), dtype=float32), tf.Tensor(-0.0143182445, shape=(), dtype=float32), tf.Tensor(0.03878837, shape=(), dtype=float32), tf.Tensor(0.074402735, shape=(), dtype=float32), tf.Tensor(0.021511186, shape=(), dtype=float32), tf.Tensor(0.01120709, shape=(), dtype=float32), tf.Tensor(-0.0055965506, shape=(), dtype=float32), tf.Tensor(-0.03297675, shape=(), dtype=float32), tf.Tensor(-0.017122181, shape=(), dtype=float32), tf.Tensor(0.023215504, shape=(), dtype=float32), tf.Tensor(0.0023373598, shape=(), dtype=float32), tf.Tensor(-0.020384455, shape=(), dtype=float32), tf.Tensor(-0.017145824, shape=(), dtype=float32), tf.Tensor(-0.04054259, shape=(), dtype=float32), tf.Tensor(0.0015166264, shape=(), dtype=float32), tf.Tensor(0.009659941, shape=(), dtype=float32), tf.Tensor(-0.011240177, shape=(), dtype=float32), tf.Tensor(-0.13036893, shape=(), dtype=float32), tf.Tensor(0.016040368, shape=(), dtype=float32), tf.Tensor(-0.017487042, shape=(), dtype=float32), tf.Tensor(-0.00030466286, shape=(), dtype=float32), tf.Tensor(0.05042587, shape=(), dtype=float32), tf.Tensor(0.021934887, shape=(), dtype=float32), tf.Tensor(0.021242432, shape=(), dtype=float32), tf.Tensor(0.023614423, shape=(), dtype=float32), tf.Tensor(0.11073799, shape=(), dtype=float32), tf.Tensor(-0.035394892, shape=(), dtype=float32), tf.Tensor(0.03901027, shape=(), dtype=float32), tf.Tensor(0.007175722, shape=(), dtype=float32), tf.Tensor(0.013500772, shape=(), dtype=float32), tf.Tensor(-0.06808823, shape=(), dtype=float32), tf.Tensor(-0.016812729, shape=(), dtype=float32), tf.Tensor(-0.02199846, shape=(), dtype=float32), tf.Tensor(-0.034011316, shape=(), dtype=float32), tf.Tensor(0.006890839, shape=(), dtype=float32), tf.Tensor(-0.0077404124, shape=(), dtype=float32), tf.Tensor(-0.016576642, shape=(), dtype=float32), tf.Tensor(-0.011995097, shape=(), dtype=float32), tf.Tensor(-0.03620341, shape=(), dtype=float32), tf.Tensor(0.050961435, shape=(), dtype=float32), tf.Tensor(-4.2990498e-05, shape=(), dtype=float32), tf.Tensor(0.034567915, shape=(), dtype=float32), tf.Tensor(-0.00025924182, shape=(), dtype=float32), tf.Tensor(0.016785251, shape=(), dtype=float32), tf.Tensor(0.0970638, shape=(), dtype=float32), tf.Tensor(0.042462207, shape=(), dtype=float32), tf.Tensor(-0.0085408, shape=(), dtype=float32), tf.Tensor(0.0313212, shape=(), dtype=float32), tf.Tensor(-0.009317408, shape=(), dtype=float32), tf.Tensor(0.06608054, shape=(), dtype=float32), tf.Tensor(-0.009573651, shape=(), dtype=float32), tf.Tensor(-0.002945682, shape=(), dtype=float32), tf.Tensor(-0.023405321, shape=(), dtype=float32), tf.Tensor(-0.03322177, shape=(), dtype=float32), tf.Tensor(-0.0069239573, shape=(), dtype=float32), tf.Tensor(-0.04210521, shape=(), dtype=float32), tf.Tensor(0.011191868, shape=(), dtype=float32), tf.Tensor(-0.04007502, shape=(), dtype=float32), tf.Tensor(-0.041769423, shape=(), dtype=float32), tf.Tensor(-0.036197573, shape=(), dtype=float32), tf.Tensor(0.005983288, shape=(), dtype=float32), tf.Tensor(-0.056538347, shape=(), dtype=float32), tf.Tensor(-0.065433554, shape=(), dtype=float32), tf.Tensor(0.0010896284, shape=(), dtype=float32), tf.Tensor(0.016883992, shape=(), dtype=float32), tf.Tensor(0.010104156, shape=(), dtype=float32), tf.Tensor(-0.017128741, shape=(), dtype=float32), tf.Tensor(-0.06581602, shape=(), dtype=float32), tf.Tensor(0.040816627, shape=(), dtype=float32), tf.Tensor(0.0073410994, shape=(), dtype=float32), tf.Tensor(-0.0074424613, shape=(), dtype=float32), tf.Tensor(-0.057583623, shape=(), dtype=float32), tf.Tensor(0.006783987, shape=(), dtype=float32), tf.Tensor(-0.020749316, shape=(), dtype=float32), tf.Tensor(-0.01439232, shape=(), dtype=float32), tf.Tensor(-0.0389859, shape=(), dtype=float32), tf.Tensor(0.10855435, shape=(), dtype=float32), tf.Tensor(0.052365422, shape=(), dtype=float32), tf.Tensor(-0.055226468, shape=(), dtype=float32), tf.Tensor(0.015788274, shape=(), dtype=float32), tf.Tensor(-0.04391703, shape=(), dtype=float32), tf.Tensor(-0.05178745, shape=(), dtype=float32), tf.Tensor(0.043569602, shape=(), dtype=float32), tf.Tensor(-0.07885928, shape=(), dtype=float32), ...)  \n",
       "300080          (tf.Tensor(0.06873965, shape=(), dtype=float32), tf.Tensor(0.011879233, shape=(), dtype=float32), tf.Tensor(0.009017263, shape=(), dtype=float32), tf.Tensor(-0.018235395, shape=(), dtype=float32), tf.Tensor(-0.07004947, shape=(), dtype=float32), tf.Tensor(-0.02003451, shape=(), dtype=float32), tf.Tensor(0.07945702, shape=(), dtype=float32), tf.Tensor(-0.048830975, shape=(), dtype=float32), tf.Tensor(-0.024501288, shape=(), dtype=float32), tf.Tensor(-0.02304068, shape=(), dtype=float32), tf.Tensor(-0.003356253, shape=(), dtype=float32), tf.Tensor(0.06627553, shape=(), dtype=float32), tf.Tensor(0.011041261, shape=(), dtype=float32), tf.Tensor(0.04809818, shape=(), dtype=float32), tf.Tensor(-0.09761175, shape=(), dtype=float32), tf.Tensor(-0.04054614, shape=(), dtype=float32), tf.Tensor(0.059656307, shape=(), dtype=float32), tf.Tensor(-0.059224084, shape=(), dtype=float32), tf.Tensor(0.026933756, shape=(), dtype=float32), tf.Tensor(-0.026851019, shape=(), dtype=float32), tf.Tensor(0.026299762, shape=(), dtype=float32), tf.Tensor(0.0060137818, shape=(), dtype=float32), tf.Tensor(0.011821149, shape=(), dtype=float32), tf.Tensor(0.041414645, shape=(), dtype=float32), tf.Tensor(-0.025367305, shape=(), dtype=float32), tf.Tensor(-0.006416064, shape=(), dtype=float32), tf.Tensor(-0.010305901, shape=(), dtype=float32), tf.Tensor(-0.004629364, shape=(), dtype=float32), tf.Tensor(0.006930405, shape=(), dtype=float32), tf.Tensor(0.090073176, shape=(), dtype=float32), tf.Tensor(0.03630198, shape=(), dtype=float32), tf.Tensor(0.028755702, shape=(), dtype=float32), tf.Tensor(-0.007877648, shape=(), dtype=float32), tf.Tensor(0.025796896, shape=(), dtype=float32), tf.Tensor(-0.062035393, shape=(), dtype=float32), tf.Tensor(-0.0046287896, shape=(), dtype=float32), tf.Tensor(0.014067956, shape=(), dtype=float32), tf.Tensor(-0.012977073, shape=(), dtype=float32), tf.Tensor(0.0029820544, shape=(), dtype=float32), tf.Tensor(-0.09998014, shape=(), dtype=float32), tf.Tensor(-0.045328774, shape=(), dtype=float32), tf.Tensor(0.0052359262, shape=(), dtype=float32), tf.Tensor(-0.044200826, shape=(), dtype=float32), tf.Tensor(-0.03938994, shape=(), dtype=float32), tf.Tensor(0.04806144, shape=(), dtype=float32), tf.Tensor(-0.014352055, shape=(), dtype=float32), tf.Tensor(-0.022317957, shape=(), dtype=float32), tf.Tensor(0.071658514, shape=(), dtype=float32), tf.Tensor(0.03830682, shape=(), dtype=float32), tf.Tensor(-0.03129089, shape=(), dtype=float32), tf.Tensor(-0.03567564, shape=(), dtype=float32), tf.Tensor(0.03654407, shape=(), dtype=float32), tf.Tensor(0.04357404, shape=(), dtype=float32), tf.Tensor(-0.027244585, shape=(), dtype=float32), tf.Tensor(0.04282825, shape=(), dtype=float32), tf.Tensor(-0.013412595, shape=(), dtype=float32), tf.Tensor(0.038874693, shape=(), dtype=float32), tf.Tensor(0.05319179, shape=(), dtype=float32), tf.Tensor(-0.0023268433, shape=(), dtype=float32), tf.Tensor(0.01715461, shape=(), dtype=float32), tf.Tensor(-0.013560631, shape=(), dtype=float32), tf.Tensor(-0.040419366, shape=(), dtype=float32), tf.Tensor(-0.006181546, shape=(), dtype=float32), tf.Tensor(0.042837176, shape=(), dtype=float32), tf.Tensor(-0.0055621546, shape=(), dtype=float32), tf.Tensor(0.0043764277, shape=(), dtype=float32), tf.Tensor(0.009759368, shape=(), dtype=float32), tf.Tensor(-0.015365781, shape=(), dtype=float32), tf.Tensor(-0.019109767, shape=(), dtype=float32), tf.Tensor(-0.06670105, shape=(), dtype=float32), tf.Tensor(-0.058795642, shape=(), dtype=float32), tf.Tensor(0.029145777, shape=(), dtype=float32), tf.Tensor(0.020474376, shape=(), dtype=float32), tf.Tensor(-0.0052547404, shape=(), dtype=float32), tf.Tensor(-0.010090535, shape=(), dtype=float32), tf.Tensor(0.07101111, shape=(), dtype=float32), tf.Tensor(0.045605715, shape=(), dtype=float32), tf.Tensor(-0.02296545, shape=(), dtype=float32), tf.Tensor(-0.021467282, shape=(), dtype=float32), tf.Tensor(-0.071915306, shape=(), dtype=float32), tf.Tensor(0.008871779, shape=(), dtype=float32), tf.Tensor(0.01678863, shape=(), dtype=float32), tf.Tensor(-0.016889704, shape=(), dtype=float32), tf.Tensor(0.039395362, shape=(), dtype=float32), tf.Tensor(-0.035669666, shape=(), dtype=float32), tf.Tensor(0.005358715, shape=(), dtype=float32), tf.Tensor(0.016329113, shape=(), dtype=float32), tf.Tensor(-0.026122803, shape=(), dtype=float32), tf.Tensor(0.026374765, shape=(), dtype=float32), tf.Tensor(0.01715521, shape=(), dtype=float32), tf.Tensor(0.008343227, shape=(), dtype=float32), tf.Tensor(-0.014655218, shape=(), dtype=float32), tf.Tensor(-0.019896802, shape=(), dtype=float32), tf.Tensor(0.09332796, shape=(), dtype=float32), tf.Tensor(-0.0077383136, shape=(), dtype=float32), tf.Tensor(-0.020521123, shape=(), dtype=float32), tf.Tensor(-8.6800566e-05, shape=(), dtype=float32), tf.Tensor(0.0622379, shape=(), dtype=float32), tf.Tensor(0.039992284, shape=(), dtype=float32), tf.Tensor(-0.067760274, shape=(), dtype=float32), ...)  \n",
       "300081                (tf.Tensor(0.023907334, shape=(), dtype=float32), tf.Tensor(0.052315816, shape=(), dtype=float32), tf.Tensor(-0.051633414, shape=(), dtype=float32), tf.Tensor(0.0077471714, shape=(), dtype=float32), tf.Tensor(0.03970935, shape=(), dtype=float32), tf.Tensor(-0.013736675, shape=(), dtype=float32), tf.Tensor(-0.012582323, shape=(), dtype=float32), tf.Tensor(-0.024585912, shape=(), dtype=float32), tf.Tensor(-0.014852009, shape=(), dtype=float32), tf.Tensor(-0.055619143, shape=(), dtype=float32), tf.Tensor(0.008263253, shape=(), dtype=float32), tf.Tensor(-0.0047130096, shape=(), dtype=float32), tf.Tensor(-0.032742828, shape=(), dtype=float32), tf.Tensor(0.05743263, shape=(), dtype=float32), tf.Tensor(-0.01613141, shape=(), dtype=float32), tf.Tensor(-0.034406934, shape=(), dtype=float32), tf.Tensor(0.038817033, shape=(), dtype=float32), tf.Tensor(0.019072218, shape=(), dtype=float32), tf.Tensor(0.011669539, shape=(), dtype=float32), tf.Tensor(-0.090077706, shape=(), dtype=float32), tf.Tensor(0.0014555468, shape=(), dtype=float32), tf.Tensor(0.005513809, shape=(), dtype=float32), tf.Tensor(-0.030070733, shape=(), dtype=float32), tf.Tensor(-0.010374811, shape=(), dtype=float32), tf.Tensor(0.039355654, shape=(), dtype=float32), tf.Tensor(-0.037163842, shape=(), dtype=float32), tf.Tensor(0.02772324, shape=(), dtype=float32), tf.Tensor(-0.02353395, shape=(), dtype=float32), tf.Tensor(0.028946167, shape=(), dtype=float32), tf.Tensor(0.017719546, shape=(), dtype=float32), tf.Tensor(0.0018260363, shape=(), dtype=float32), tf.Tensor(0.00028033735, shape=(), dtype=float32), tf.Tensor(-0.052647278, shape=(), dtype=float32), tf.Tensor(0.0756154, shape=(), dtype=float32), tf.Tensor(0.020653268, shape=(), dtype=float32), tf.Tensor(-0.06249076, shape=(), dtype=float32), tf.Tensor(-0.036298044, shape=(), dtype=float32), tf.Tensor(0.011630775, shape=(), dtype=float32), tf.Tensor(0.068808146, shape=(), dtype=float32), tf.Tensor(-0.040713735, shape=(), dtype=float32), tf.Tensor(-0.01589599, shape=(), dtype=float32), tf.Tensor(0.022711592, shape=(), dtype=float32), tf.Tensor(0.010543281, shape=(), dtype=float32), tf.Tensor(-0.016693002, shape=(), dtype=float32), tf.Tensor(-0.0031882452, shape=(), dtype=float32), tf.Tensor(0.005573282, shape=(), dtype=float32), tf.Tensor(-0.038671143, shape=(), dtype=float32), tf.Tensor(-0.0023794014, shape=(), dtype=float32), tf.Tensor(-0.052908592, shape=(), dtype=float32), tf.Tensor(0.03668991, shape=(), dtype=float32), tf.Tensor(-0.013172233, shape=(), dtype=float32), tf.Tensor(-0.0239118, shape=(), dtype=float32), tf.Tensor(0.031588282, shape=(), dtype=float32), tf.Tensor(0.06316237, shape=(), dtype=float32), tf.Tensor(-0.017381046, shape=(), dtype=float32), tf.Tensor(0.04683652, shape=(), dtype=float32), tf.Tensor(0.027002355, shape=(), dtype=float32), tf.Tensor(0.041966658, shape=(), dtype=float32), tf.Tensor(0.095873885, shape=(), dtype=float32), tf.Tensor(0.06987687, shape=(), dtype=float32), tf.Tensor(-0.048182003, shape=(), dtype=float32), tf.Tensor(-0.02679722, shape=(), dtype=float32), tf.Tensor(0.01095338, shape=(), dtype=float32), tf.Tensor(0.0124590155, shape=(), dtype=float32), tf.Tensor(-0.03715376, shape=(), dtype=float32), tf.Tensor(0.011673558, shape=(), dtype=float32), tf.Tensor(0.025146147, shape=(), dtype=float32), tf.Tensor(0.005816766, shape=(), dtype=float32), tf.Tensor(-0.011063083, shape=(), dtype=float32), tf.Tensor(-0.0067656143, shape=(), dtype=float32), tf.Tensor(-0.07596833, shape=(), dtype=float32), tf.Tensor(0.07020325, shape=(), dtype=float32), tf.Tensor(0.07631031, shape=(), dtype=float32), tf.Tensor(0.050751124, shape=(), dtype=float32), tf.Tensor(-0.014694677, shape=(), dtype=float32), tf.Tensor(-0.009856238, shape=(), dtype=float32), tf.Tensor(0.0121922465, shape=(), dtype=float32), tf.Tensor(0.05429827, shape=(), dtype=float32), tf.Tensor(-0.02645367, shape=(), dtype=float32), tf.Tensor(0.013764614, shape=(), dtype=float32), tf.Tensor(-0.08916231, shape=(), dtype=float32), tf.Tensor(0.05923999, shape=(), dtype=float32), tf.Tensor(-0.042189505, shape=(), dtype=float32), tf.Tensor(-0.06420892, shape=(), dtype=float32), tf.Tensor(-0.01678818, shape=(), dtype=float32), tf.Tensor(0.015626308, shape=(), dtype=float32), tf.Tensor(-0.015270449, shape=(), dtype=float32), tf.Tensor(-0.030836768, shape=(), dtype=float32), tf.Tensor(-0.015567685, shape=(), dtype=float32), tf.Tensor(-0.0399256, shape=(), dtype=float32), tf.Tensor(0.05246063, shape=(), dtype=float32), tf.Tensor(-0.04651136, shape=(), dtype=float32), tf.Tensor(0.031117456, shape=(), dtype=float32), tf.Tensor(-0.023506302, shape=(), dtype=float32), tf.Tensor(-0.07396878, shape=(), dtype=float32), tf.Tensor(0.040847186, shape=(), dtype=float32), tf.Tensor(-0.0040281336, shape=(), dtype=float32), tf.Tensor(0.052818257, shape=(), dtype=float32), tf.Tensor(0.019324802, shape=(), dtype=float32), tf.Tensor(0.08227731, shape=(), dtype=float32), ...)  \n",
       "300082                        (tf.Tensor(0.016358314, shape=(), dtype=float32), tf.Tensor(-0.055671506, shape=(), dtype=float32), tf.Tensor(-0.029637381, shape=(), dtype=float32), tf.Tensor(-0.056957223, shape=(), dtype=float32), tf.Tensor(0.052265424, shape=(), dtype=float32), tf.Tensor(-0.008139698, shape=(), dtype=float32), tf.Tensor(-0.030265389, shape=(), dtype=float32), tf.Tensor(0.027607191, shape=(), dtype=float32), tf.Tensor(0.032367367, shape=(), dtype=float32), tf.Tensor(0.04106494, shape=(), dtype=float32), tf.Tensor(0.014060544, shape=(), dtype=float32), tf.Tensor(-0.0021166166, shape=(), dtype=float32), tf.Tensor(-0.08100254, shape=(), dtype=float32), tf.Tensor(0.030571435, shape=(), dtype=float32), tf.Tensor(0.035408, shape=(), dtype=float32), tf.Tensor(-0.03086979, shape=(), dtype=float32), tf.Tensor(0.04915643, shape=(), dtype=float32), tf.Tensor(-0.022131559, shape=(), dtype=float32), tf.Tensor(0.034481965, shape=(), dtype=float32), tf.Tensor(-0.0038810829, shape=(), dtype=float32), tf.Tensor(-0.028350623, shape=(), dtype=float32), tf.Tensor(-0.02234467, shape=(), dtype=float32), tf.Tensor(0.037224375, shape=(), dtype=float32), tf.Tensor(0.000751521, shape=(), dtype=float32), tf.Tensor(-0.042181917, shape=(), dtype=float32), tf.Tensor(-0.056799695, shape=(), dtype=float32), tf.Tensor(0.0901877, shape=(), dtype=float32), tf.Tensor(0.039094556, shape=(), dtype=float32), tf.Tensor(-0.011283439, shape=(), dtype=float32), tf.Tensor(-0.044154435, shape=(), dtype=float32), tf.Tensor(-0.04247438, shape=(), dtype=float32), tf.Tensor(0.0069788494, shape=(), dtype=float32), tf.Tensor(-0.030908998, shape=(), dtype=float32), tf.Tensor(-0.04946975, shape=(), dtype=float32), tf.Tensor(-0.046985492, shape=(), dtype=float32), tf.Tensor(-0.058152452, shape=(), dtype=float32), tf.Tensor(-0.039360806, shape=(), dtype=float32), tf.Tensor(-0.018504063, shape=(), dtype=float32), tf.Tensor(0.09645569, shape=(), dtype=float32), tf.Tensor(0.02572451, shape=(), dtype=float32), tf.Tensor(0.108570255, shape=(), dtype=float32), tf.Tensor(0.03272362, shape=(), dtype=float32), tf.Tensor(-0.020669617, shape=(), dtype=float32), tf.Tensor(0.03645965, shape=(), dtype=float32), tf.Tensor(0.006018922, shape=(), dtype=float32), tf.Tensor(0.034145925, shape=(), dtype=float32), tf.Tensor(-0.03645899, shape=(), dtype=float32), tf.Tensor(-0.06911441, shape=(), dtype=float32), tf.Tensor(0.0016972882, shape=(), dtype=float32), tf.Tensor(0.1351007, shape=(), dtype=float32), tf.Tensor(0.03229023, shape=(), dtype=float32), tf.Tensor(-0.039639413, shape=(), dtype=float32), tf.Tensor(0.029782634, shape=(), dtype=float32), tf.Tensor(0.04943118, shape=(), dtype=float32), tf.Tensor(0.014977691, shape=(), dtype=float32), tf.Tensor(0.047863957, shape=(), dtype=float32), tf.Tensor(-0.041000806, shape=(), dtype=float32), tf.Tensor(-0.041627176, shape=(), dtype=float32), tf.Tensor(0.02688666, shape=(), dtype=float32), tf.Tensor(0.04648368, shape=(), dtype=float32), tf.Tensor(0.04881043, shape=(), dtype=float32), tf.Tensor(-0.014024911, shape=(), dtype=float32), tf.Tensor(0.040062383, shape=(), dtype=float32), tf.Tensor(0.048952438, shape=(), dtype=float32), tf.Tensor(0.026217744, shape=(), dtype=float32), tf.Tensor(0.015916433, shape=(), dtype=float32), tf.Tensor(0.004321039, shape=(), dtype=float32), tf.Tensor(-0.020352382, shape=(), dtype=float32), tf.Tensor(0.029310487, shape=(), dtype=float32), tf.Tensor(-0.038597338, shape=(), dtype=float32), tf.Tensor(-0.036106978, shape=(), dtype=float32), tf.Tensor(-0.032617886, shape=(), dtype=float32), tf.Tensor(-0.00010633847, shape=(), dtype=float32), tf.Tensor(0.058943268, shape=(), dtype=float32), tf.Tensor(-0.031699147, shape=(), dtype=float32), tf.Tensor(0.024221651, shape=(), dtype=float32), tf.Tensor(-0.027113795, shape=(), dtype=float32), tf.Tensor(0.041020654, shape=(), dtype=float32), tf.Tensor(0.024591466, shape=(), dtype=float32), tf.Tensor(-0.062298473, shape=(), dtype=float32), tf.Tensor(0.01828891, shape=(), dtype=float32), tf.Tensor(0.044462077, shape=(), dtype=float32), tf.Tensor(0.010415857, shape=(), dtype=float32), tf.Tensor(-0.069450855, shape=(), dtype=float32), tf.Tensor(-0.008549134, shape=(), dtype=float32), tf.Tensor(0.020632012, shape=(), dtype=float32), tf.Tensor(-0.010715748, shape=(), dtype=float32), tf.Tensor(0.0095858, shape=(), dtype=float32), tf.Tensor(-0.05704763, shape=(), dtype=float32), tf.Tensor(-0.0030989805, shape=(), dtype=float32), tf.Tensor(-0.01156347, shape=(), dtype=float32), tf.Tensor(-0.018763319, shape=(), dtype=float32), tf.Tensor(0.04045568, shape=(), dtype=float32), tf.Tensor(0.0009379599, shape=(), dtype=float32), tf.Tensor(-0.07353291, shape=(), dtype=float32), tf.Tensor(0.08545972, shape=(), dtype=float32), tf.Tensor(0.09345397, shape=(), dtype=float32), tf.Tensor(-0.01154483, shape=(), dtype=float32), tf.Tensor(0.01290133, shape=(), dtype=float32), tf.Tensor(-0.022994723, shape=(), dtype=float32), ...)  \n",
       "300083        (tf.Tensor(-0.07129504, shape=(), dtype=float32), tf.Tensor(0.05181775, shape=(), dtype=float32), tf.Tensor(-0.02489603, shape=(), dtype=float32), tf.Tensor(-0.023506913, shape=(), dtype=float32), tf.Tensor(0.01281394, shape=(), dtype=float32), tf.Tensor(0.08956977, shape=(), dtype=float32), tf.Tensor(0.037577093, shape=(), dtype=float32), tf.Tensor(0.051022176, shape=(), dtype=float32), tf.Tensor(-0.05336184, shape=(), dtype=float32), tf.Tensor(0.054678135, shape=(), dtype=float32), tf.Tensor(0.015588774, shape=(), dtype=float32), tf.Tensor(0.012392177, shape=(), dtype=float32), tf.Tensor(-0.008402479, shape=(), dtype=float32), tf.Tensor(0.023921112, shape=(), dtype=float32), tf.Tensor(0.01599294, shape=(), dtype=float32), tf.Tensor(0.012201126, shape=(), dtype=float32), tf.Tensor(0.12451981, shape=(), dtype=float32), tf.Tensor(-0.00078481203, shape=(), dtype=float32), tf.Tensor(0.021262959, shape=(), dtype=float32), tf.Tensor(-0.009026773, shape=(), dtype=float32), tf.Tensor(-0.005325233, shape=(), dtype=float32), tf.Tensor(-0.0049861637, shape=(), dtype=float32), tf.Tensor(0.070501365, shape=(), dtype=float32), tf.Tensor(-0.0037700974, shape=(), dtype=float32), tf.Tensor(0.034333933, shape=(), dtype=float32), tf.Tensor(0.02061151, shape=(), dtype=float32), tf.Tensor(-0.03962808, shape=(), dtype=float32), tf.Tensor(-0.013045015, shape=(), dtype=float32), tf.Tensor(0.02984808, shape=(), dtype=float32), tf.Tensor(-0.006684023, shape=(), dtype=float32), tf.Tensor(-0.10910741, shape=(), dtype=float32), tf.Tensor(0.049963016, shape=(), dtype=float32), tf.Tensor(0.010760641, shape=(), dtype=float32), tf.Tensor(0.0079353135, shape=(), dtype=float32), tf.Tensor(0.0815682, shape=(), dtype=float32), tf.Tensor(-0.014976661, shape=(), dtype=float32), tf.Tensor(0.018694662, shape=(), dtype=float32), tf.Tensor(-0.00011071347, shape=(), dtype=float32), tf.Tensor(0.015297066, shape=(), dtype=float32), tf.Tensor(0.017451882, shape=(), dtype=float32), tf.Tensor(-0.026891377, shape=(), dtype=float32), tf.Tensor(-0.060761075, shape=(), dtype=float32), tf.Tensor(-0.03561407, shape=(), dtype=float32), tf.Tensor(-0.060945064, shape=(), dtype=float32), tf.Tensor(0.08811272, shape=(), dtype=float32), tf.Tensor(0.03339591, shape=(), dtype=float32), tf.Tensor(0.046238407, shape=(), dtype=float32), tf.Tensor(-0.033360608, shape=(), dtype=float32), tf.Tensor(0.076256156, shape=(), dtype=float32), tf.Tensor(0.04862837, shape=(), dtype=float32), tf.Tensor(0.041696865, shape=(), dtype=float32), tf.Tensor(-0.051193308, shape=(), dtype=float32), tf.Tensor(-0.03433186, shape=(), dtype=float32), tf.Tensor(0.03920537, shape=(), dtype=float32), tf.Tensor(-0.05263466, shape=(), dtype=float32), tf.Tensor(-0.022765348, shape=(), dtype=float32), tf.Tensor(0.012549712, shape=(), dtype=float32), tf.Tensor(-0.042384088, shape=(), dtype=float32), tf.Tensor(-0.0037845653, shape=(), dtype=float32), tf.Tensor(0.038839225, shape=(), dtype=float32), tf.Tensor(-0.010468014, shape=(), dtype=float32), tf.Tensor(-0.013704077, shape=(), dtype=float32), tf.Tensor(-0.013660341, shape=(), dtype=float32), tf.Tensor(0.0622114, shape=(), dtype=float32), tf.Tensor(0.06779714, shape=(), dtype=float32), tf.Tensor(-0.016206617, shape=(), dtype=float32), tf.Tensor(-0.06212112, shape=(), dtype=float32), tf.Tensor(-0.021148777, shape=(), dtype=float32), tf.Tensor(-0.025203573, shape=(), dtype=float32), tf.Tensor(-0.0026878437, shape=(), dtype=float32), tf.Tensor(-0.038446885, shape=(), dtype=float32), tf.Tensor(-0.030847214, shape=(), dtype=float32), tf.Tensor(0.021817302, shape=(), dtype=float32), tf.Tensor(0.0075747157, shape=(), dtype=float32), tf.Tensor(0.016853029, shape=(), dtype=float32), tf.Tensor(0.01750912, shape=(), dtype=float32), tf.Tensor(0.024633944, shape=(), dtype=float32), tf.Tensor(-0.013615712, shape=(), dtype=float32), tf.Tensor(-0.0030897972, shape=(), dtype=float32), tf.Tensor(-0.008850621, shape=(), dtype=float32), tf.Tensor(-0.012036455, shape=(), dtype=float32), tf.Tensor(-0.022593455, shape=(), dtype=float32), tf.Tensor(-0.053547565, shape=(), dtype=float32), tf.Tensor(-0.080733314, shape=(), dtype=float32), tf.Tensor(0.0066507645, shape=(), dtype=float32), tf.Tensor(0.020736605, shape=(), dtype=float32), tf.Tensor(0.012811926, shape=(), dtype=float32), tf.Tensor(0.03420647, shape=(), dtype=float32), tf.Tensor(-0.030252958, shape=(), dtype=float32), tf.Tensor(-0.056493968, shape=(), dtype=float32), tf.Tensor(-0.017451148, shape=(), dtype=float32), tf.Tensor(-0.015234978, shape=(), dtype=float32), tf.Tensor(0.020664064, shape=(), dtype=float32), tf.Tensor(0.12766692, shape=(), dtype=float32), tf.Tensor(-0.07579299, shape=(), dtype=float32), tf.Tensor(-0.00022272844, shape=(), dtype=float32), tf.Tensor(0.028153209, shape=(), dtype=float32), tf.Tensor(-0.04204758, shape=(), dtype=float32), tf.Tensor(-0.10946826, shape=(), dtype=float32), tf.Tensor(-0.031043392, shape=(), dtype=float32), ...)  \n",
       "\n",
       "[288783 rows x 4 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trope_examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "d7fb9b1f-0927-4784-bd3c-f43190a23aca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.95907015"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_similarity(trope_examples['embedding'].iloc[0], trope_examples['embedding'].iloc[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "5a2a6066-728a-419e-b8d2-17d561e61efc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.45953935"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_similarity(trope_examples['embedding'].iloc[0], trope_examples['embedding'].iloc[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "72ac845f-fb4c-42a7-a774-75969940ecc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "trope_examples.to_csv('trope_examples_embeddings.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f981e042-535c-43d1-a62b-b628bd98dbdd",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}