Replies: 6 comments 2 replies
-
You can use |
Beta Was this translation helpful? Give feedback.
-
How to use
|
Beta Was this translation helpful? Give feedback.
-
@yangsp5 I think @mdouze meant |
Beta Was this translation helpful? Give feedback.
-
IndexScalarQuantizer support float16? |
Beta Was this translation helpful? Give feedback.
-
acebook.com/profile.php?id=100054340704401 |
Beta Was this translation helpful? Give feedback.
-
The conversion from fp32 to fp16 is done internally. |
Beta Was this translation helpful? Give feedback.
-
Summary
need
IndexFlatIP
supportfloat16
when the number of vector is very very large, such as
1e10
. UseIndexFlatIP
offloat32
is too expensive, maybefloat16
is much fastterBeta Was this translation helpful? Give feedback.
All reactions