Skip to main content

​
Added

  • Added new flags to the train command: arch, model path, parallelism.

​
Changed

  • Add GPU node limits based on the selected architecture through the arch flag.