Skip to content

v.1.3.7

Latest

Choose a tag to compare

@LovelyBuggies LovelyBuggies released this 06 Mar 22:26
· 7 commits to main since this release

Changelog

  • Remove the redundant sampling hyperparameters in algorithms.
  • Allow multi-gpu training with MP (see test results at PR #62).

Notes: Currently, we limit the scope of parallel training via model parallel for research use (4x7/15B models should be fine). Ray-based distributed training for a larger scope is unnecessary and has not been developed yet.