Custom layout parser Model detect slow

[prasum]

Hi Team,

I have trained custom layout parser detection 2 model and tried model.detect on cpu. It takes around 4s on 2 cores+8GB RAM. Currently 1 image is passed to model.detect. Could you please let me know if batch of images can be passed to model.detect and how the model inference time can be improved? Request your guidance on the same.

[hmiche]

Hello sir do you have any ressource how to train a custom model ?

[prasum]

Hi @hmiche sorry for the late reply you can have a look at https://github.com/Layout-Parser/layout-model-training repository.