UX-Decoder / Segment-Everything-Everywhere-All-At-Once

[NeurIPS 2023] Official implementation of the paper "Segment Everything Everywhere All at Once"
Apache License 2.0
4.41k stars 408 forks source link

How to change the size of the training set? #136

Open EricZavier opened 8 months ago

EricZavier commented 8 months ago

Thank you very much for your outstanding work,How to change the size of the training set? The training time is too long according to the dataset in the Redme document. I want to conduct small sample training, but I am missing any image code from the dataset mentioned in Redme and an error is reported.