jackboyla/gone_and_growned_my_own_dataset

---
license: apache-2.0
size_categories:
- 1K<n<10K
dataset_info:
features:
- name: id
dtype: int64
- name: text
sequence: string
- name: tokenized_text
sequence:
sequence: string
- name: model_name
dtype: string
- name: instruction
dtype: string
- name: ents
list:
list:
- name: head
sequence: string
- name: tail
sequence: string
- name: generation
sequence: string
- name: ner
sequence:
sequence: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 1148647044
num_examples: 26688
download_size: 141700312
dataset_size: 1148647044
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
tags:
- synthetic
---

## Dataset structure

The examples have the following structure per configuration:

<details><summary> Configuration: default </summary><hr>

```json
{
"ents": {
"head": [
"41",
"42",
"PRODUCT",
"IPhone"
],
"tail": [
"0",
"5",
"DATE",
"This time of the year"
]
},
"generation": " NO_RELATION",
"instruction": "You are a fantastic relation extraction model who only outputs valid JSON.\nExtract the relation between the given entities using the context in the below text. If no relation exists, use the label \"NO_RELATION\".\nONLY RETURN THE RELATION LABEL.\nPay VERY close attention to which entity is the head and tail; this dictates the direction of the relationship.\n\nEntities: {\u0027head\u0027: [\u002741\u0027, \u002742\u0027, \u0027PRODUCT\u0027, \u0027IPhone\u0027], \u0027tail\u0027: [\u00270\u0027, \u00275\u0027, \u0027DATE\u0027, \u0027This time of the year\u0027]}\n\nText: This time of the year...\n...my heart sings, my energy level pumps up and I feel inspired.\nDoes that happen to you too?\nJust thought I would pop in to share some quick IPhone photos\nthat I have taken in the last few days.\nI love playing with my lilacs.\nIf you have followed along these last four years,\nyou know that our lilacs are family blooms...\nsome were my grandmother\u0027s, Mr. Flea\u0027s grandmother\u0027s\nand the deep double French were my Mom\u0027s.\nThe dogwoods this past week were stunning.\nWe have just one pink...the rest are white.\nAnd...of course, I love using my vintage, rustic treasures\nto hold the all the beautiful blossoms\nwe are lucky enough to have on the property.\nMy fascination with Instagram continues,\nbut my son Dan will be happy to know\nthat I have been using my Canon big girl camera again\nand now that we have replaced my 8 year old laptop\n(may she rest in peace)\nwith a new one that is fast and has lots of storage...\nI hope to blog a bit more often as well.\nWishing you a lovely spring!\nI can\u0027t wait to get back up to the lake!",
"model_name": "mistralai/Mistral-7B-Instruct-v0.2",
"text": "This time of the year...\n...my heart sings, my energy level pumps up and I feel inspired.\nDoes that happen to you too?\nJust thought I would pop in to share some quick IPhone photos\nthat I have taken in the last few days.\nI love playing with my lilacs.\nIf you have followed along these last four years,\nyou know that our lilacs are family blooms...\nsome were my grandmother\u0027s, Mr. Flea\u0027s grandmother\u0027s\nand the deep double French were my Mom\u0027s.\nThe dogwoods this past week were stunning.\nWe have just one pink...the rest are white.\nAnd...of course, I love using my vintage, rustic treasures\nto hold the all the beautiful blossoms\nwe are lucky enough to have on the property.\nMy fascination with Instagram continues,\nbut my son Dan will be happy to know\nthat I have been using my Canon big girl camera again\nand now that we have replaced my 8 year old laptop\n(may she rest in peace)\nwith a new one that is fast and has lots of storage...\nI hope to blog a bit more often as well.\nWishing you a lovely spring!\nI can\u0027t wait to get back up to the lake!",
"tokenized_text": [
"This",
"time",
"of",
"the",
"year",
"...",
"\n",
"...",
"my",
"heart",
"sings",
",",
"my",
"energy",
"level",
"pumps",
"up",
"and",
"I",
"feel",
"inspired",
".",
"\n",
"Does",
"that",
"happen",
"to",
"you",
"too",
"?",
"\n",
"Just",
"thought",
"I",
"would",
"pop",
"in",
"to",
"share",
"some",
"quick",
"IPhone",
"photos",
"\n",
"that",
"I",
"have",
"taken",
"in",
"the",
"last",
"few",
"days",
".",
"\n",
"I",
"love",
"playing",
"with",
"my",
"lilacs",
".",
"\n",
"If",
"you",
"have",
"followed",
"along",
"these",
"last",
"four",
"years",
",",
"\n",
"you",
"know",
"that",
"our",
"lilacs",
"are",
"family",
"blooms",
"...",
"\n",
"some",
"were",
"my",
"grandmother",
"\u0027s",
",",
"Mr.",
"Flea",
"\u0027s",
"grandmother",
"\u0027s",
"\n",
"and",
"the",
"deep",
"double",
"French",
"were",
"my",
"Mom",
"\u0027s",
".",
"\n",
"The",
"dogwoods",
"this",
"past",
"week",
"were",
"stunning",
".",
"\n",
"We",
"have",
"just",
"one",
"pink",
"...",
"the",
"rest",
"are",
"white",
".",
"\n",
"And",
"...",
"of",
"course",
",",
"I",
"love",
"using",
"my",
"vintage",
",",
"rustic",
"treasures",
"\n",
"to",
"hold",
"the",
"all",
"the",
"beautiful",
"blossoms",
"\n",
"we",
"are",
"lucky",
"enough",
"to",
"have",
"on",
"the",
"property",
".",
"\n",
"My",
"fascination",
"with",
"Instagram",
"continues",
",",
"\n",
"but",
"my",
"son",
"Dan",
"will",
"be",
"happy",
"to",
"know",
"\n",
"that",
"I",
"have",
"been",
"using",
"my",
"Canon",
"big",
"girl",
"camera",
"again",
"\n",
"and",
"now",
"that",
"we",
"have",
"replaced",
"my",
"8",
"year",
"old",
"laptop",
"\n",
"(",
"may",
"she",
"rest",
"in",
"peace",
")",
"\n",
"with",
"a",
"new",
"one",
"that",
"is",
"fast",
"and",
"has",
"lots",
"of",
"storage",
"...",
"\n",
"I",
"hope",
"to",
"blog",
"a",
"bit",
"more",
"often",
"as",
"well",
".",
"\n",
"Wishing",
"you",
"a",
"lovely",
"spring",
"!",
"\n",
"I",
"ca",
"n\u0027t",
"wait",
"to",
"get",
"back",
"up",
"to",
"the",
"lake",
"!"
]
}
```

This subset can be loaded as:

```python
from datasets import load_dataset

ds = load_dataset("jackboyla/gone_and_growned_my_own_dataset", "default")
```

Or simply as it follows, since there's only one configuration and is named `default`:

```python
from datasets import load_dataset

ds = load_dataset("jackboyla/gone_and_growned_my_own_dataset")
```

</details>

原始信息汇总

数据集概述

许可证

• Apache 2.0

数据规模

• 1K < n < 10K

数据集信息

• 特征：
  • id：int64
  • text：字符串序列
  • tokenized_text：字符串序列的序列
  • model_name：字符串
  • instruction：字符串
  • ents：列表的列表
    • head：字符串序列
    • tail：字符串序列
  • generation：字符串序列
  • ner：字符串序列的序列
  • __index_level_0__：int64

数据分割

• 训练集：
  • 字节数：1148647044
  • 样本数：26688

数据大小

• 下载大小：141700312
• 数据集大小：1148647044

配置

• 默认配置：
  • 数据文件：
    • 分割：训练
    • 路径：data/train-*

标签

• 合成数据

数据结构示例（默认配置）

json { "ents": { "head": ["41", "42", "PRODUCT", "IPhone"], "tail": ["0", "5", "DATE", "This time of the year"] }, "generation": "NO_RELATION", "instruction": "You are a fantastic relation extraction model who only outputs valid JSON. Extract the relation between the given entities using the context in the below text. If no relation exists, use the label "NO_RELATION". ONLY RETURN THE RELATION LABEL. Pay VERY close attention to which entity is the head and tail; this dictates the direction of the relationship.

Entities: {head: [41, 42, PRODUCT, IPhone], tail: [0, 5, DATE, This time of the year]}

Text: This time of the year... ...my heart sings, my energy level pumps up and I feel inspired. Does that happen to you too? Just thought I would pop in to share some quick IPhone photos that I have taken in the last few days. I love playing with my lilacs. If you have followed along these last four years, you know that our lilacs are family blooms... some were my grandmothers, Mr. Fleas grandmothers and the deep double French were my Moms. The dogwoods this past week were stunning. We have just one pink...the rest are white. And...of course, I love using my vintage, rustic treasures to hold the all the beautiful blossoms we are lucky enough to have on the property. My fascination with Instagram continues, but my son Dan will be happy to know that I have been using my Canon big girl camera again and now that we have replaced my 8 year old laptop (may she rest in peace) with a new one that is fast and has lots of storage... I hope to blog a bit more often as well. Wishing you a lovely spring! I cant wait to get back up to the lake!", "model_name": "mistralai/Mistral-7B-Instruct-v0.2", "text": "This time of the year... ...my heart sings, my energy level pumps up and I feel inspired. Does that happen to you too? Just thought I would pop in to share some quick IPhone photos that I have taken in the last few days. I love playing with my lilacs. If you have followed along these last four years, you know that our lilacs are family blooms... some were my grandmothers, Mr. Fleas grandmothers and the deep double French were my Moms. The dogwoods this past week were stunning. We have just one pink...the rest are white. And...of course, I love using my vintage, rustic treasures to hold the all the beautiful blossoms we are lucky enough to have on the property. My fascination with Instagram continues, but my son Dan will be happy to know that I have been using my Canon big girl camera again and now that we have replaced my 8 year old laptop (may she rest in peace) with a new one that is fast and has lots of storage... I hope to blog a bit more often as well. Wishing you a lovely spring! I cant wait to get back up to the lake!", "tokenized_text": ["This", "time", "of", "the", "year", "...", " ", "...", "my", "heart", "sings", ",", "my", "energy", "level", "pumps", "up", "and", "I", "feel", "inspired", ".", " ", "Does", "that", "happen", "to", "you", "too", "?", " ", "Just", "thought", "I", "would", "pop", "in", "to", "share", "some", "quick", "IPhone", "photos", " ", "that", "I", "have", "taken", "in", "the", "last", "few", "days", ".", " ", "I", "love", "playing", "with", "my", "lilacs", ".", " ", "If", "you", "have", "followed", "along", "these", "last", "four", "years", ",", " ", "you", "know", "that", "our", "lilacs", "are", "family", "blooms", "...", " ", "some", "were", "my", "grandmothers", ",", "Mr.", "Flea", "grandmothers", " ", "and", "the", "deep", "double", "French", "were", "my", "Moms", ".", " ", "The", "dogwoods", "this", "past", "week", "were", "stunning", ".", " ", "We", "have", "just", "one", "pink", "...", "the", "rest", "are", "white", ".", " ", "And", "...", "of", "course", ",", "I", "love", "using", "my", "vintage", ",", "rustic", "treasures", " ", "to", "hold", "the", "all", "the", "beautiful", "blossoms", " ", "we", "are", "lucky", "enough", "to", "have", "on", "the", "property", ".", " ", "My", "fascination", "with", "Instagram", "continues", ",", " ", "but", "my", "son", "Dan", "will", "be", "happy", "to", "know", " ", "that", "I", "have", "been", "using", "my", "Canon", "big", "girl", "camera", "again", " ", "and", "now", "that", "we", "have", "replaced", "my", "8", "year", "old", "laptop", " ", "(", "may", "she", "rest", "in", "peace", ")", " ", "with", "a", "new", "one", "that", "is", "fast", "and", "has", "lots", "of", "storage", "...", " ", "I", "hope", "to", "blog", "a", "bit", "more", "often", "as", "well", ".", " ", "Wishing", "you", "a", "lovely", "spring", "!", " ", "I", "ca", "nt", "wait", "to", "get", "back", "up", "to", "the", "lake", "!"] }

加载数据集

python from datasets import load_dataset

ds = load_dataset("jackboyla/gone_and_growned_my_own_dataset", "default")

或简化为： python from datasets import load_dataset

ds = load_dataset("jackboyla/gone_and_growned_my_own_dataset")