Duplicate Datasets?
#1
by
mrfakename
- opened
Hi,
I saw the model card mentioned it was trained on both Open Orca and mlabonne's ChatML DPO dataset, however mlabonne's ChatML DPO dataset is simply Open Orca converted to the ChatML format.
Was this model trained to accept multiple chat formats by using the Open Orca dataset and then the ChatML dataset?
Thanks!