qidian

bbox, landmark = self.model.detect(cons_img, threshold=0.1, scale=1.0) problem

1

self.model = insightface.model_zoo.get_model('retinaface_r50_v1') # ./.insightface self.model.prepare(ctx_id=self.gpu_id, nms=0.4) bbox, landmark = self.model.detect(cons_img, threshold=0.1, scale=1.0) I want to use this code to perform detection, but the latest version of the `insightface` library...

Hello, could you please share your trained pth file?