Skip to content

v0.0.2

Compare
Choose a tag to compare
@PanQiWei PanQiWei released this 23 Apr 12:06
· 701 commits to main since this release
4f84d21
  • added eval_tasks module to support evaluate model's performance on predefined down-stream tasks before and after quantization
  • fixed some bugs when using LLaMa model
  • fixed some bugs when using models that required position_ids