Testing time for the inference stage

[Ziyu Jiang]

Hi,

I am very interested in this work and I am trying to test its performance.
I successfully run the code, but the speed I got is about 1s/frame on the V100 GPU, which is far slower than what the paper reported. Do you have any idea why it would be like this.

Thanks.

[lzozo95]

HI,

I'm so sorry that this comment is not about your query. I want to ask you a question.
How can you download the normal weight file and how to use it? Isn't there any error?

Thanks.

[Ziyu Jiang]

HI,

I'm so sorry that this comment is not about your query. I want to ask you a question.
How can you download the normal weight file and how to use it? Isn't there any error?

Thanks.

I just download without unzipping. The torch.load can directly read it.

[pingjun18-li]

HI,
I'm so sorry that this comment is not about your query. I want to ask you a question.
How can you download the normal weight file and how to use it? Isn't there any error?
Thanks.

I just download without unzipping. The torch.load can directly read it.

I have download the weight, but when i load it , it print the error as below:

import torch
torch.load("/data/VOS/GraphMemVOS-master/workspace_STM_alpha/main_runfile_graph_memory.pth.tar")
Traceback (most recent call last):
File "", line 1, in
File "/opt/conda/lib/python3.6/site-packages/torch/serialization.py", line 368, in load
return _load(f, map_location, pickle_module)
File "/opt/conda/lib/python3.6/site-packages/torch/serialization.py", line 542, in _load
result = unpickler.load()
ModuleNotFoundError: No module named 'utils.stats'
or can anyone who can read the weight to save it to nomally .pth but npt .pth.tar? thank u very much