Do you fine-tuned benchmark dataset?
#7
by linkai-dl - opened
As stated in the readme, the following datasets were used.
openbookqa
sciq
Open-Orca/OpenOrca
metaeval/ScienceQA_text_only
GAIR/lima
Hello @linkai-dl .
As stated here, we did not use any benchmark datasets, including training or testing ones.
wonhosong changed discussion status to closed