Update train_mixtral_8x7b.yaml

FlagOpen / FlagScale

FlagScale is a large model toolkit based on open-sourced projects.

Other

132 stars 40 forks source link

Update train_mixtral_8x7b.yaml #189

Closed shenzhu1993 closed 1 month ago

shenzhu1993 commented 1 month ago

This PR, based on the the Mixtral-8x7B paper and our tests , fixes some bugs that were present in the train_mixtral_8x7b.yaml file and modifies some unnecessary parameters.