Task
Other
P
bigscience/P3
Other · bigscience· 105.4K
["apache-2.0"]
466 GB
Dataset Card for P3 Dataset Summary P3 (Public 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… See the full description on the dataset page: https://huggingface.co/datasets/bigscience/P3.
# download instantly
mlforge datasets pull bigscience/P3
mlforge datasets pull bigscience/P3
Dataset details
Language
en
License
["apache-2.0"]
Size
466 GB
Creator
bigscience
Downloads
105.4K
Source
huggingface_datasets
Updated
2024-03-04
About bigscience/P3
--- annotationscreators: - crowdsourced - expert-generated language: - en license: - apache-2.0 multilinguality: - monolingual sizecategories: - 100M<n<1B taskcategories: - other prettyname: P3 datasetinfo: - configname: adversarialqadbertanswerthefollowingq features: - name: inputs sequence: int32 - name: inputspretokenized dtype: string - name: targets sequence: int32 - name: targetspretokenized dtype: string splits: - name: train numbytes: 18313753 numexamples: 10000 - name: validation numbytes: 1791034 numexamples: 1000 downloadsize: 6288641 datasetsize: 20104787 - configname: adversarialqadbertbasedon features: - name: inputs sequence: int32 - name: inputspretokenized dtype: string - name: targets sequence: int32 - name: targetspretokenized dtype: string splits: - name: train numbytes: 17580553 numexamples: 10000 - name: validation numbytes: 1717566 numexamples: 1000 downloadsize: 6206744 datasetsize: 19298119 - configname: adversarialqadbertgeneratequestion features: - name: inputs sequence: int32 - name: inputspretokenized dtype: string - name: targets sequence: int32 - name: targetspretokenized dtype: string splits: - name: train numbytes: 18552810 numexamples: 10000 - name: validation numbytes: 1824231 numexamples: 1000 - name: test numbytes: 1954952 numexamples: 1000 downloadsize: 5882604 datasetsize: 22331993 - configname: adversarialqadbertquestioncontextanswer features: - name: inputs seque