You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'd like to raise an issue regarding the rounding implementations of tfloat32 and bfloat16. Why does tfloat32 use round-to-nearest, while bfloat16 uses round-to-nearest-even?
shanliang1992
changed the title
[QST]
[QST] Inconsistency in Rounding Implementations: Round-to-Nearest for TFloat32 vs. Round-to-Nearest-Even for BFloat16
Oct 30, 2024
This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.
I'd like to raise an issue regarding the rounding implementations of tfloat32 and bfloat16. Why does tfloat32 use round-to-nearest, while bfloat16 uses round-to-nearest-even?
tfloat32
bfloat16
The text was updated successfully, but these errors were encountered: