Giters

OpenNMT / CTranslate2

Fast inference engine for Transformer models

Home Page:https://opennmt.net/CTranslate2

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Ctranslate2 Pypi exceeds limit 20GB

minhthuc2502 opened this issue · comments

commented

Currently, the Pypi hub of Ctranslate2 exceeds 20GB, the release 4.3.0 is missing for python 3.8 and 3.9. Request is sent to pypi to upgrade the limit. Some points to review:

  • Delete old versions (2.x)
  • Consider to deactivate FlashAttention in python package (which is quite heavy but does not make any big improvement in the performance compared with CT2 standard MHA)