Inference for a new image with the pretrained model

[Shiying99]

Hello,

Thank you for the amazing work. I am new here and would like to test some pretrained model for scene graph generation like the DEMO using my laptop. It is a windows laptop and was successfully installed all the necessary packages using conda and pip and the datasets.

However, I don't have slurm on my machine. So I wonder if it is possible to run the inference model with config file on the windows machine to recreate the same output as the Demo API.

Few questions:

1. Where can I change the input dataset for inferencing instead of using the preset one?

2. When I run following command direct on the terminal without slurm, I got an error from mmdet packge, how to solve it:
   python tools/test.py configs/psgtr/psgtr_r50_psg.py work_dirs/psgtr_r50_e60/epoch_60.pth --out work_dirs/psgtr_r50_e60/epoch_60_extra.pkl --show-dir work_dirs/psgtr_r50_e60

creating index...
index created!
load checkpoint from local path: work_dirs/psgtr_r50_e60/epoch_60.pth
[ ] 0/2177, elapsed: 0s, ETA:Traceback (most recent call last):
File "tools/test.py", line 244, in
main()
File "tools/test.py", line 207, in main
outputs = single_gpu_test(model, data_loader, args.show, args.show_dir,
File "C:\Users\ShiyingLi\anaconda3\envs\mmdet\lib\site-packages\mmdet\apis\test.py", line 52, in single_gpu_test
model.module.show_result(
File "C:\Users\ShiyingLi\anaconda3\envs\mmdet\lib\site-packages\mmdet\models\detectors\base.py", line 320, in show_result
bboxes = np.vstack(bbox_result)
File "<array_function internals>", line 200, in vstack
File "C:\Users\ShiyingLi\anaconda3\envs\mmdet\lib\site-packages\numpy\core\shape_base.py", line 293, in vstack
arrs = atleast_2d(*tup)
File "<array_function internals>", line 200, in atleast_2d
File "C:\Users\ShiyingLi\anaconda3\envs\mmdet\lib\site-packages\numpy\core\shape_base.py", line 121, in atleast_2d
ary = asanyarray(ary)
ValueError: setting an array element with a sequence. The requested array would exceed the maximum number of dimension of 32.

@chenxwh @Jingkang50

Really appreciate if you reply to the questions. Thank you so much.

[Jingkang50]

Thank you for your interest in our project!
I assume the error comes from some version issue? Does your cuda and mmdet version match our readme requirement?