Datasets:

Languages:
English
ArXiv:
License:
stellaathena commited on
Commit
52ed156
1 Parent(s): 26efd77

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +30 -13
README.md CHANGED
@@ -5,13 +5,12 @@ language_creators:
5
  - found
6
  language:
7
  - en
8
- license:
9
- - other
10
  multilinguality:
11
  - monolingual
12
- pretty_name: The Pile
13
  size_categories:
14
- - unknown
15
  source_datasets:
16
  - original
17
  task_categories:
@@ -50,6 +49,8 @@ paperswithcode_id: the-pile
50
  - [Citation Information](#citation-information)
51
  - [Contributions](#contributions)
52
 
 
 
53
  ## Dataset Description
54
 
55
  - **Homepage:** https://pile.eleuther.ai/
@@ -57,6 +58,7 @@ paperswithcode_id: the-pile
57
  - **Paper:** [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://arxiv.org/abs/2101.00027)
58
  - **Leaderboard:**
59
  - **Point of Contact:** [EleutherAI](mailto:[email protected])
 
60
 
61
  ### Dataset Summary
62
 
@@ -83,6 +85,8 @@ This dataset is in English (`EN`).
83
  'text': 'It is done, and submitted. You can play “Survival of the Tastiest” on Android, and on the web. Playing on...'
84
  }
85
  ```
 
 
86
 
87
  #### enron_emails
88
  ```
@@ -166,6 +170,8 @@ This dataset is in English (`EN`).
166
  }
167
  ```
168
 
 
 
169
  ### Data Fields
170
 
171
  #### all
@@ -174,6 +180,9 @@ This dataset is in English (`EN`).
174
  - `meta` (dict): Metadata of the data instance with keys:
175
  - pile_set_name: Name of the subset.
176
 
 
 
 
177
  #### enron_emails
178
 
179
  - `text` (str): Text.
@@ -229,6 +238,8 @@ This dataset is in English (`EN`).
229
 
230
  The "all" configuration is composed of 3 splits: train, validation and test.
231
 
 
 
232
  ## Dataset Creation
233
 
234
  ### Curation Rationale
@@ -277,7 +288,7 @@ The "all" configuration is composed of 3 splits: train, validation and test.
277
 
278
  ### Dataset Curators
279
 
280
- [More Information Needed]
281
 
282
  ### Licensing Information
283
 
@@ -287,16 +298,22 @@ Please refer to the specific license depending on the subset you use:
287
  ### Citation Information
288
 
289
  ```
290
- @misc{gao2020pile,
291
- title={The Pile: An 800GB Dataset of Diverse Text for Language Modeling},
292
- author={Leo Gao and Stella Biderman and Sid Black and Laurence Golding and Travis Hoppe and Charles Foster and Jason Phang and Horace He and Anish Thite and Noa Nabeshima and Shawn Presser and Connor Leahy},
293
- year={2020},
294
- eprint={2101.00027},
295
- archivePrefix={arXiv},
296
- primaryClass={cs.CL}
 
 
 
 
 
 
297
  }
298
  ```
299
 
300
  ### Contributions
301
 
302
- Thanks to [@github-username](https://github.com/<github-username>) for adding this dataset.
 
5
  - found
6
  language:
7
  - en
8
+ license: other
 
9
  multilinguality:
10
  - monolingual
11
+ pretty_name: the Pile
12
  size_categories:
13
+ - 100B<n<1T
14
  source_datasets:
15
  - original
16
  task_categories:
 
49
  - [Citation Information](#citation-information)
50
  - [Contributions](#contributions)
51
 
52
+ This model card is a work in progress. Please also see [our datasheet](https://arxiv.org/abs/2201.07311) for more detailed info.
53
+
54
  ## Dataset Description
55
 
56
  - **Homepage:** https://pile.eleuther.ai/
 
58
  - **Paper:** [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://arxiv.org/abs/2101.00027)
59
  - **Leaderboard:**
60
  - **Point of Contact:** [EleutherAI](mailto:[email protected])
61
+ - **Datasheet:** [Datasheet for the Pile](https://arxiv.org/abs/2201.07311)
62
 
63
  ### Dataset Summary
64
 
 
85
  'text': 'It is done, and submitted. You can play “Survival of the Tastiest” on Android, and on the web. Playing on...'
86
  }
87
  ```
88
+ <details>
89
+ <summary>Expand to see individual components</summary>
90
 
91
  #### enron_emails
92
  ```
 
170
  }
171
  ```
172
 
173
+ </details>
174
+
175
  ### Data Fields
176
 
177
  #### all
 
180
  - `meta` (dict): Metadata of the data instance with keys:
181
  - pile_set_name: Name of the subset.
182
 
183
+ <details>
184
+ <summary>Expand to see individual components</summary>
185
+
186
  #### enron_emails
187
 
188
  - `text` (str): Text.
 
238
 
239
  The "all" configuration is composed of 3 splits: train, validation and test.
240
 
241
+ </details>
242
+
243
  ## Dataset Creation
244
 
245
  ### Curation Rationale
 
288
 
289
  ### Dataset Curators
290
 
291
+ This dataset was primarily curated by Leo Gao and Stella Biderman, with assistance from other authors of the Pile paper.
292
 
293
  ### Licensing Information
294
 
 
298
  ### Citation Information
299
 
300
  ```
301
+ @article{gao2020pile,
302
+ title={The {P}ile: An 800{GB} dataset of diverse text for language modeling},
303
+ author={Gao, Leo and Biderman, Stella and Black, Sid and Golding, Laurence and Hoppe, Travis and Foster, Charles and Phang, Jason and He, Horace and Thite, Anish and Nabeshima, Noa and others},
304
+ journal={arXiv preprint arXiv:2101.00027},
305
+ year={2020}
306
+ }
307
+
308
+
309
+ @article{biderman2022datasheet,
310
+ title={Datasheet for the pile},
311
+ author={Biderman, Stella and Bicheno, Kieran and Gao, Leo},
312
+ journal={arXiv preprint arXiv:2201.07311},
313
+ year={2022}
314
  }
315
  ```
316
 
317
  ### Contributions
318
 
319
+ Thanks to [@github-username](https://github.com/<github-username>) for adding this dataset.