PromptSource
@promptsourceSource account
Imported from https://github.com/bigscience-workshop/promptsource. Is this yours? Claim it from any prompt page.
- Followers
- 0 followers
- Following
- 0 following
- Prompts
- 430 prompts
- Joined
- Joined September 2026
Prompts
openai_humaneval: function body generation
PromptSource template for the openai_humaneval dataset.
0 stars
nq_open: question_answer
Plain Question
0 stars
nq_open: formal_description
Copied from the dataset description.
0 stars
nq_open: first_person_context
Ask a question in first person
0 stars
nlu_evaluation_data: intent_to_query
PromptSource template for the nlu_evaluation_data dataset.
0 stars
nlu_evaluation_data: what_does_user_want
PromptSource template for the nlu_evaluation_data dataset.
0 stars
nlu_evaluation_data: classify_intent
PromptSource template for the nlu_evaluation_data dataset.
0 stars
newspop: topic_no_choice_in_template
PromptSource template for the newspop dataset.
0 stars
newspop: before_text_question_no_choice
PromptSource template for the newspop dataset.
0 stars
newspop: after_text_question_no_choice
PromptSource template for the newspop dataset.
0 stars
neural_code_search/evaluation_dataset: generate code given a description
PromptSource template for the neural_code_search/evaluation_dataset dataset.
0 stars
neural_code_search/evaluation_dataset: generate a description given code
PromptSource template for the neural_code_search/evaluation_dataset dataset.
0 stars
multi_x_science_sum: generate_abstract_from_related_work
PromptSource template for the multi_x_science_sum dataset.
0 stars
multi_x_science_sum: generate_related_work_from_abstract
PromptSource template for the multi_x_science_sum dataset.
0 stars
multi_nli: based on the previous passage
Adapted from the BoolQ prompts in Schick & Schütze 2021.
0 stars
multi_nli: MNLI crowdsource
Adapted from Williams et al. 2018's instructions to crowdsourcing workers.
0 stars
multi_nli: GPT-3 style
Same as reported in Figure G7 of the GPT-3 paper, except that there is no task identifying tokens like "anli R1: ".
0 stars
mocha: score_candidate_with_question
Similarity measure between candidate and reference answers (in a regression manner)
0 stars
mocha: score_candidate_no_ques_no_context_interrogative
Similarity measure between candidate and reference answers (in a regression manner)
0 stars
mocha: score_candidate_no_ques_no_context_affirmative
PromptSource template for the mocha dataset.
0 stars