Segmentation fault (core dumped)

[KlZOnmyway]

Hi, I'm trying to use your Infer for a single MP4 video cmd. However, no matter which model I chose, e.g., FasterVQA, Fast-VQA-M, it raises error: Segmentation fault (core dumped). I tried to debug it. It seems the problem happened when I tried to put model onto device. However, if I use CPU to run it, the bash will not respond anymore.

My torch version is 2.2.0+cu118. System: Ubuntu. GPU: V100 32GB

[sunwei925]

the same problem

[MOLYHECI]

the same.

[Black-Knight875]

Also happens to me. After testing, I found that the error is related to a conflict between decord and torch.
To resolve this problem, you just need to place all

import torch

before

import decord

You can see the files that need to be modified using the git grep decord command. I modified the following files:

• vqa.py
• fastvqa/datasets/.ipynb_checkpoints/basic_datasets-checkpoint.py
• fastvqa/datasets/.ipynb_checkpoints/fusion_datasets-checkpoint.py
• fastvqa/datasets/basic_datasets.py
• fastvqa/datasets/fusion_datasets.py

Versions:

torch: 1.11.0
torchvision: 0.12.0
decord: 0.6.0

I also tried the following versions, and the issue can be resolved in the same way:

torch: 2.4.1
torchvision: 0.19.1
decord: 0.6.0