qa-validation
COVID-QA-unique-context-test-10-percent-validation-10-percent
Dataset Card for "COVID-QA-unique-context-test-10-percent-validation-10-percent"
More Information needed
downstream_validation_qatruthful_qa-validation-german_q_n_atrivia_qa_validation_mc
TriviaQA Shortcut
Overview
This repository contains a multiple-choice reformulation of the TriviaQA dataset benchmark, created as part of our work on reducing the computational cost of evaluating large language models (LLMs) during (pre-)training.
Each item contains an answers column, where the first item is always the original correct answer from TriviaQA. The remaining answer options are distractors generated by Meta-Llama-3.1-70B-Instruct-GPTQ-INT4. For details… See the full description on the dataset page: https://huggingface.co/datasets/fraunhofer-iis/trivia_qa_validation_mc.qa_validation_qwenCOVID-QA-train-80-test-10-validation-10
Dataset Card for "COVID-QA-train-80-test-10-validation-10"
More Information needed