Adaptation to HF Multimodal Processor
#1
by
mlinmg
- opened
No description provided.
mlinmg
changed pull request title from
Adaptation to HF Multimodal Processt
to Adaptation to HF Multimodal Processor
I think this will be very helpful both for pipeline unification and for implementation in other inference engine.
I have specifically changed a bit hte logic by removing the preprocessing logic form the model and I've ported it to the HF preprocessor component
mlinmg
changed pull request status to
open
One point of impvement could be to add the video preprocessing to the Processor