torchtune icon indicating copy to clipboard operation
torchtune copied to clipboard

Change default dataset in DPO configs to use HH-RLHF dataset

Open SalmanMohammadi opened this issue 5 months ago • 0 comments

cc @RdoubleA we'll have to re-benchmark here

SalmanMohammadi avatar Sep 20 '24 11:09 SalmanMohammadi