Inference on LazyConfig

[eksis77]

Hi,
I try to train with LazyConfig_train_net.py. ( not yaml )
I successed the training phase with my custom dataset, but I don't know how to infer from this model.

In following code, I loaded the trained model.
but Inference does't work...

please, let me know how to infer from an image

========================================
config_file = "configs/cascade_mask_rcnn_mvitv2_t_3x.py"
cfg = LazyConfig.load(config_file)

model = instantiate(cfg.model)
model.to(cfg.train.device)
model = create_ddp_model(model)
DetectionCheckpointer(model).load("output/model_final.pth")
model.eval()

im = cv2.imread('/data/test/001.jpg')
outputs = model(im)