Need suggestions to finetune
#25
by
Venkatesh4342
- opened
if we want to finetune the omni parser on new element identification:
1.) Then should we just need to finetune our custom yolo model on sepecific element and when ever omniparser fails should we just load our finetuned model and work as it is.
2.) Do we need to finetuned the text description model( florence-2 or Blip-2) also for the new element ?