I want library users to be able to construct supported scene-text recognizer models by public model id…

I want library users to be able to construct supported scene-text recognizer models by public model id…: a task in MiMo-V2.6-RL-harbor-code: MiMo-V2.6-RL Code (Harbor) (Harbor dataset). For create model('parseq', pretrained=False) , return a PARSeq model initialized from the shared defaults plus…