There are many aspect to "integrate". What is the actual need?
a
Amit Gelber
03/16/2023, 1:28 PM
serve models with ray & bento.
couldn't find it in the docs, but did in the code
j
Jiang
03/16/2023, 1:35 PM
Can I ask a deeper question? Why do you need to use both BentoML and Ray for serving? Is it because of migration?
a
Amit Gelber
03/16/2023, 2:59 PM
sure, it's for scaling batch inference
Amit Gelber
03/16/2023, 2:59 PM
same reason spark integration exist
s
Sean
03/16/2023, 11:48 PM
@Amit Gelber, Ray integration is currently a work-in-progress, not officially announced. The integration is going to be with RayServe. Is that what you had in mind?