VictorSanh HF staff commited on
Commit
a03bea0
1 Parent(s): e37c8bf

update README + gitignore

Browse files
Files changed (2) hide show
  1. .gitignore +114 -0
  2. README.md +41 -71
.gitignore ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Byte-compiled / optimized / DLL files
2
+ __pycache__/
3
+ *.py[cod]
4
+ *$py.class
5
+ # C extensions
6
+ *.so
7
+ # Distribution / packaging
8
+ .Python
9
+ build/
10
+ develop-eggs/
11
+ dist/
12
+ downloads/
13
+ eggs/
14
+ .eggs/
15
+ lib/
16
+ lib64/
17
+ parts/
18
+ sdist/
19
+ var/
20
+ wheels/
21
+ share/python-wheels/
22
+ *.egg-info/
23
+ .installed.cfg
24
+ *.egg
25
+ MANIFEST
26
+ # PyInstaller
27
+ # Usually these files are written by a python script from a template
28
+ # before PyInstaller builds the exe, so as to inject date/other infos into it.
29
+ *.manifest
30
+ *.spec
31
+ # Installer logs
32
+ pip-log.txt
33
+ pip-delete-this-directory.txt
34
+ # Unit test / coverage reports
35
+ htmlcov/
36
+ .tox/
37
+ .nox/
38
+ .coverage
39
+ .coverage.*
40
+ .cache
41
+ nosetests.xml
42
+ coverage.xml
43
+ *.cover
44
+ *.py,cover
45
+ .hypothesis/
46
+ .pytest_cache/
47
+ cover/
48
+ # Translations
49
+ *.mo
50
+ *.pot
51
+ # Django stuff:
52
+ *.log
53
+ local_settings.py
54
+ db.sqlite3
55
+ db.sqlite3-journal
56
+ # Flask stuff:
57
+ instance/
58
+ .webassets-cache
59
+ # Scrapy stuff:
60
+ .scrapy
61
+ # Sphinx documentation
62
+ docs/_build/
63
+ # PyBuilder
64
+ .pybuilder/
65
+ target/
66
+ # Jupyter Notebook
67
+ .ipynb_checkpoints
68
+ # IPython
69
+ profile_default/
70
+ ipython_config.py
71
+ # pyenv
72
+ # For a library or package, you might want to ignore these files since the code is
73
+ # intended to run in multiple environments; otherwise, check them in:
74
+ # .python-version
75
+ # pipenv
76
+ # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
77
+ # However, in case of collaboration, if having platform-specific dependencies or dependencies
78
+ # having no cross-platform support, pipenv may install dependencies that don't work, or not
79
+ # install all needed dependencies.
80
+ #Pipfile.lock
81
+ # PEP 582; used by e.g. github.com/David-OConnor/pyflow
82
+ __pypackages__/
83
+ # Celery stuff
84
+ celerybeat-schedule
85
+ celerybeat.pid
86
+ # SageMath parsed files
87
+ *.sage.py
88
+ # Environments
89
+ .env
90
+ .venv
91
+ env/
92
+ venv/
93
+ ENV/
94
+ env.bak/
95
+ venv.bak/
96
+ # Spyder project settings
97
+ .spyderproject
98
+ .spyproject
99
+ # Rope project settings
100
+ .ropeproject
101
+ # mkdocs documentation
102
+ /site
103
+ # mypy
104
+ .mypy_cache/
105
+ .dmypy.json
106
+ dmypy.json
107
+ # Pyre type checker
108
+ .pyre/
109
+ # pytype static type analyzer
110
+ .pytype/
111
+ # Cython debug symbols
112
+ cython_debug/
113
+ # Datasets locks
114
+ *.lock
README.md CHANGED
@@ -1,7 +1,19 @@
1
  ---
2
  YAML tags:
3
- - copy-paste the tags obtained with the tagging app: https://github.com/huggingface/datasets-tagging
4
- - WIP!
 
 
 
 
 
 
 
 
 
 
 
 
5
  ---
6
 
7
  # Dataset Card for P3
@@ -35,16 +47,16 @@ YAML tags:
35
 
36
  - **Homepage:** https://bigscience.huggingface.co/promptsource
37
  - **Repository:** https://github.com/bigscience-workshop/promptsource/
38
- - **Paper:** TBA
39
  - **Point of Contact:** Victor Sanh (victor@huggingface.co)
40
 
41
  ### Dataset Summary
42
 
43
- P3 is a collection of prompted English datasets covering a diverse set of NLP tasks. A prompt is the combination of an input template and a target template. The templates are functions mapping a data example into natural language for the input and target sequences. For example, in the case of an NLI dataset, the data example would include fields for *Premise, Hypothesis, Label*. An input template would be *If {Premise} is true, is it also true that {Hypothesis}?*, whereas a target template can be defined with the label choices *Choices[label]*. Here *Choices* is prompt-specific metadata that consists of the options *yes, maybe, no* corresponding to *label* being entailment (0), neutral (1) or contradiction (2).
44
 
45
- Prompts are collected using [Promptsource](https://github.com/bigscience-workshop/promptsource), an interface to interactively write prompts on datasets, and collect prompt-specific metadata such as evaluation metrics. As of October 13th, there are 2'000 prompts collected for 270+ data(sub)sets. The collection of prompts is publicly available on [Promptsource](https://github.com/bigscience-workshop/promptsource).
46
 
47
- To train [T0*](https://huggingface.co/bigscience/T0pp_11B), we used a subset of the prompts available in Promptsource (see details [here](https://huggingface.co/bigscience/T0pp_11B#training-data)). However, some of the prompts use `random.choice`, a method that selects uniformly at random an option in a list of valid possibilities. For reproducibility purposes, we release the collection of prompted examples used to train T0*. **The data available here are the materialized version of the prompted datasets used in [Multi-task enables task zero-shot generalization](TODO) which represent only a subset datasets for which there is at least one prompt on Promptsource.**
48
 
49
  ### Supported Tasks and Leaderboards
50
 
@@ -60,29 +72,35 @@ The data in P3 are in English (BCP-47 `en`).
60
 
61
  An example of "train" looks as follows:
62
  ```bash
 
63
  ```
64
 
65
- To check all the prompted examples, you can use [Promptsource hosted tool](http://bigscience.huggingface.co/promptsource) and choose the `Prompted dataset viewer` mode in the left panel.
66
 
67
 
68
  ### Data Fields
69
 
70
  The data fields are the same among all splits:
71
- - `input_text`: the natural language input fed to the model
72
- - `target_text`: the natural language target that the model has to generate
73
- - `tokenized_input`: the tokenized `input_text` with T5's tokenizer
74
- - `tokenized_target`: the tokenized `target_text` with T5's tokenizer
 
 
 
 
75
 
76
  ### Data Splits
77
 
78
  |Data(sub)set|Split|Number of examples|
79
  |-|-|-|
 
80
 
81
  ## Dataset Creation
82
 
83
  ### Curation Rationale
84
 
85
- P3 relies on the Hugging Face Dataset library. Any public dataset in the Datasets library can be prompted. We select the datasets that have at least one subset in English and excluded datasets containing (predominantly) non-natural language examples.
86
 
87
  We conservatively decided not to prompt datasets that contain potentially harmful content (for instance, datasets built on social media content). However, we sometimes prompt datasets that are purposefully built to measure bias and fairness of trained models, and reserve these prompted datasets (the validation or test sets) for evaluation purposes.
88
 
@@ -91,40 +109,33 @@ We conservatively decided not to prompt datasets that contain potentially harmfu
91
  Here's the full list of the datasets present in the materialized version of P3:
92
  - Multiple-Choice QA
93
  - CommonsenseQA
94
- - Cosmos
95
  - DREAM
96
- - QASC
97
  - QUAIL
98
- - Quarrel
99
  - QuaRTz
100
- - SciQ
101
  - Social IQA
102
- - Wiki Hop
103
  - WiQA
 
 
 
 
 
104
  - ARC
105
- - BoolQ
106
- - Circa
107
- - MC-TACO
108
- - MultiRC
109
  - OpenBookQA
 
110
  - PIQA
111
  - RACE
 
 
112
  - Extractive QA
113
  - Adversarial QA
114
- - DuoRC
115
  - Quoref
 
116
  - ROPES
117
- - TyDiQA
118
- - CoQA
119
- - DROP
120
- - QA SRL
121
- - QuAC
122
- - ReCoRD
123
  - SQuAD v2
 
124
  - Close-book QA
125
  - Hotpot QA
126
  - Wiki QA
127
- - NQ Open
128
  - Trivia QA
129
  - Web Questions
130
  - Structure-to-text
@@ -164,15 +175,6 @@ Here's the full list of the datasets present in the materialized version of P3:
164
  - HellaSwag
165
  - Story Cloze
166
 
167
- TODO: recheck this list to match Figure 2 in final version of paper
168
-
169
- <!-- #### Initial Data Collection and Normalization
170
-
171
-
172
- #### Who are the source language producers?
173
-
174
- [More Information Needed] -->
175
-
176
  ### Annotations
177
 
178
  The prompts available in Promptsource are collected as part of BigScience, one-year long research workshop on large multilingual models and datasets. 36 contributors affiliated with 24 institutions in 8 countries participated to the prompt collection. Contributors are in majority machine learning researchers or machine learning engineers.
@@ -181,38 +183,6 @@ The main annotation guideline was that prompts needed to be grammatical and unde
181
 
182
  The full annotation given to the contributors can be found [here](https://github.com/bigscience-workshop/promptsource/blob/main/CONTRIBUTING.md). *Note to self: the link is currently being updated with the)
183
 
184
- <!-- #### Annotation process
185
-
186
- [More Information Needed]
187
-
188
- #### Who are the annotators?
189
-
190
- [More Information Needed] -->
191
-
192
- <!-- ### Personal and Sensitive Information
193
-
194
- [More Information Needed]
195
-
196
- ## Considerations for Using the Data
197
-
198
- ### Social Impact of Dataset
199
-
200
- [More Information Needed]
201
-
202
- ### Discussion of Biases
203
-
204
- [More Information Needed] -->
205
-
206
- <!-- ### Other Known Limitations
207
-
208
- [More Information Needed]
209
-
210
- ## Additional Information
211
-
212
- ### Dataset Curators
213
-
214
- [More Information Needed] -->
215
-
216
  ### Licensing Information
217
 
218
  The dataset is released under Apache 2.0.
@@ -220,7 +190,7 @@ The dataset is released under Apache 2.0.
220
  ### Citation Information
221
 
222
  ```bibtex
223
- WIP
224
  ```
225
 
226
  ### Contributions
1
  ---
2
  YAML tags:
3
+ annotations_creators:
4
+ - crowdsourced
5
+ - expert-generated
6
+ languages:
7
+ - en
8
+ licenses:
9
+ - apache-2.0
10
+ multilinguality:
11
+ - monolingual
12
+ pretty_name: P3
13
+ size_categories:
14
+ - 10M<n<100M
15
+ task_categories:
16
+ - other
17
  ---
18
 
19
  # Dataset Card for P3
47
 
48
  - **Homepage:** https://bigscience.huggingface.co/promptsource
49
  - **Repository:** https://github.com/bigscience-workshop/promptsource/
50
+ - **Paper:** TODO
51
  - **Point of Contact:** Victor Sanh (victor@huggingface.co)
52
 
53
  ### Dataset Summary
54
 
55
+ P3 (Pubic Pool of Prompts) is a collection of prompted English datasets covering a diverse set of NLP tasks. A prompt is the combination of an input template and a target template. The templates are functions mapping a data example into natural language for the input and target sequences. For example, in the case of an NLI dataset, the data example would include fields for *Premise, Hypothesis, Label*. An input template would be *If {Premise} is true, is it also true that {Hypothesis}?*, whereas a target template can be defined with the label choices *Choices[label]*. Here *Choices* is prompt-specific metadata that consists of the options *yes, maybe, no* corresponding to *label* being entailment (0), neutral (1) or contradiction (2).
56
 
57
+ Prompts are collected using [Promptsource](https://github.com/bigscience-workshop/promptsource), an interface to interactively write prompts on datasets, and collect prompt-specific metadata such as evaluation metrics. As of October 13th, there are 2'000 prompts collected for 270+ data(sub)sets. The collection of prompts of P3 is publicly available on [Promptsource](https://github.com/bigscience-workshop/promptsource).
58
 
59
+ To train [T0*](https://huggingface.co/bigscience/T0pp), we used a subset of the prompts available in Promptsource (see details [here](https://huggingface.co/bigscience/T0pp#training-data)). However, some of the prompts use `random.choice`, a method that selects uniformly at random an option in a list of valid possibilities. For reproducibility purposes, we release the collection of prompted examples used to train T0*. **The data available here are the materialized version of the prompted datasets used in [{Multitask Prompted Training Enables Zero-Shot Task Generalization](TODO) which represent only a subset of the datasets for which there is at least one prompt on Promptsource.**
60
 
61
  ### Supported Tasks and Leaderboards
62
 
72
 
73
  An example of "train" looks as follows:
74
  ```bash
75
+ TODO
76
  ```
77
 
78
+ To check all the prompted examples, you can use the [Promptsource hosted tool](http://bigscience.huggingface.co/promptsource) and choose the `Prompted dataset viewer` mode in the left panel.
79
 
80
 
81
  ### Data Fields
82
 
83
  The data fields are the same among all splits:
84
+ - `answer_choices`: the choices (in natural language) available to the model
85
+ - `inputs_pretokenized`: the natural language input fed to the model
86
+ - `targets_pretokenized`: the natural language target that the model has to generate
87
+ - `inputs`: the tokenized input with T5's tokenizer
88
+ - `targets`: the tokenized target with T5's tokenizer
89
+ - `idx`: identifier of the (example, option) in the case of rank classification
90
+ - `weight`: a weight for the example produced by seqio (always set to 1.0)
91
+ - `is_correct`: whether the (example, option) is the correct one
92
 
93
  ### Data Splits
94
 
95
  |Data(sub)set|Split|Number of examples|
96
  |-|-|-|
97
+ |WIP|WIP|WIP|
98
 
99
  ## Dataset Creation
100
 
101
  ### Curation Rationale
102
 
103
+ The Public Pool of Prompts relies on the Hugging Face Dataset library. Any public dataset in the Datasets library can be prompted. We select the datasets that have at least one subset in English and excluded datasets containing (predominantly) non-natural language examples.
104
 
105
  We conservatively decided not to prompt datasets that contain potentially harmful content (for instance, datasets built on social media content). However, we sometimes prompt datasets that are purposefully built to measure bias and fairness of trained models, and reserve these prompted datasets (the validation or test sets) for evaluation purposes.
106
 
109
  Here's the full list of the datasets present in the materialized version of P3:
110
  - Multiple-Choice QA
111
  - CommonsenseQA
 
112
  - DREAM
 
113
  - QUAIL
 
114
  - QuaRTz
 
115
  - Social IQA
 
116
  - WiQA
117
+ - Cosmos
118
+ - QASC
119
+ - Quarel
120
+ - SciQ
121
+ - Wiki Hop
122
  - ARC
 
 
 
 
123
  - OpenBookQA
124
+ - MultiRC
125
  - PIQA
126
  - RACE
127
+ - HellaSwag
128
+ - BoolQ
129
  - Extractive QA
130
  - Adversarial QA
 
131
  - Quoref
132
+ - DuoRC
133
  - ROPES
 
 
 
 
 
 
134
  - SQuAD v2
135
+ - ReCoRD
136
  - Close-book QA
137
  - Hotpot QA
138
  - Wiki QA
 
139
  - Trivia QA
140
  - Web Questions
141
  - Structure-to-text
175
  - HellaSwag
176
  - Story Cloze
177
 
 
 
 
 
 
 
 
 
 
178
  ### Annotations
179
 
180
  The prompts available in Promptsource are collected as part of BigScience, one-year long research workshop on large multilingual models and datasets. 36 contributors affiliated with 24 institutions in 8 countries participated to the prompt collection. Contributors are in majority machine learning researchers or machine learning engineers.
183
 
184
  The full annotation given to the contributors can be found [here](https://github.com/bigscience-workshop/promptsource/blob/main/CONTRIBUTING.md). *Note to self: the link is currently being updated with the)
185
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
186
  ### Licensing Information
187
 
188
  The dataset is released under Apache 2.0.
190
  ### Citation Information
191
 
192
  ```bibtex
193
+ TODO
194
  ```
195
 
196
  ### Contributions