Minimum hardware requirement for the evaluation of 720p/1080p video about dvqa

[TANYaYaPang]

Thanks for the release of DVQA.
I have run the eval.py on my workstation, which has the following specs. But somehow I always meet the error "CUDA out of memory" when evaluating 720p and 1080p videos. Only videos with the resolution of 540p or lower worked fine . I am wondering if there is any configuration I can modify to evaluate 720p/1080p videos on my workstation, or simply I just need to upgrade the hardware.

OS: Windows 10 home CPU: Intel i5-10400 RAM: 16GB GPU: RTX2060 with 6GB memory SSD: 256GB

Looking forward to your reply, Thanks a lot.

Hello, have you reproduced the DVQA code? If it is reproduced, how do you put your LIVE data set? I put all the data sets in a folder, and an error occurred: RuntimeError: The size of tensor a (35) must match the size of tensor b (109) at non-singleton dimension 2. If you know how to solve it, please reply. Looking forward to your reply, Thanks a lot.

[haiqwang]

Thanks for the release of DVQA.
I have run the eval.py on my workstation, which has the following specs. But somehow I always meet the error "CUDA out of memory" when evaluating 720p and 1080p videos. Only videos with the resolution of 540p or lower worked fine . I am wondering if there is any configuration I can modify to evaluate 720p/1080p videos on my workstation, or simply I just need to upgrade the hardware.

OS: Windows 10 home CPU: Intel i5-10400 RAM: 16GB GPU: RTX2060 with 6GB memory SSD: 256GB

Looking forward to your reply, Thanks a lot.

I have no experience about the exact memory requirements for 720p and 1080p videos. Besides upgrading hardware, one quick workaround is to reduce the number of frames per sample. The GPU memory requirement is increasing linearly with frame number.

[haiqwang]

Thanks for the release of DVQA.
I have run the eval.py on my workstation, which has the following specs. But somehow I always meet the error "CUDA out of memory" when evaluating 720p and 1080p videos. Only videos with the resolution of 540p or lower worked fine . I am wondering if there is any configuration I can modify to evaluate 720p/1080p videos on my workstation, or simply I just need to upgrade the hardware.
OS: Windows 10 home CPU: Intel i5-10400 RAM: 16GB GPU: RTX2060 with 6GB memory SSD: 256GB
Looking forward to your reply, Thanks a lot.

Hello, have you reproduced the DVQA code? If it is reproduced, how do you put your LIVE data set? I put all the data sets in a folder, and an error occurred: RuntimeError: The size of tensor a (35) must match the size of tensor b (109) at non-singleton dimension 2. If you know how to solve it, please reply. Looking forward to your reply, Thanks a lot.

A script is provided under https://github.com/Tencent/DVQA/blob/master/dataset/LIVE/prep_live_score.py. It will output a json file that could be directly used by the train or eval scripts.

Unfortunately, I am not able to figure out what happens given the limited information provided. Have you decoded the mp4 streams to YUV?

[TANYaYaPang]

Thank you for your reply. DVQA code contains the code to convert MP4 to YUV. If you enter data in MP4 format, it will be automatically converted to YUV format.

…

------------------&nbsp;原始邮件&nbsp;------------------ 发件人: "Tencent/DVQA" <[email protected]&gt;; 发送时间:&nbsp;2020年12月3日(星期四) 中午11:52 收件人:&nbsp;"Tencent/DVQA"<[email protected]&gt;; 抄送:&nbsp;"小胖子不高兴"<[email protected]&gt;;"Comment"<[email protected]&gt;; 主题:&nbsp;Re: [Tencent/DVQA] Minimum hardware requirement for the evaluation of 720p/1080p video (#13) Thanks for the release of DVQA. I have run the eval.py on my workstation, which has the following specs. But somehow I always meet the error "CUDA out of memory" when evaluating 720p and 1080p videos. Only videos with the resolution of 540p or lower worked fine . I am wondering if there is any configuration I can modify to evaluate 720p/1080p videos on my workstation, or simply I just need to upgrade the hardware. OS: Windows 10 home CPU: Intel i5-10400 RAM: 16GB GPU: RTX2060 with 6GB memory SSD: 256GB Looking forward to your reply, Thanks a lot. Hello, have you reproduced the DVQA code? If it is reproduced, how do you put your LIVE data set? I put all the data sets in a folder, and an error occurred: RuntimeError: The size of tensor a (35) must match the size of tensor b (109) at non-singleton dimension 2. If you know how to solve it, please reply. Looking forward to your reply, Thanks a lot. A script is provided under https://github.com/Tencent/DVQA/blob/master/dataset/LIVE/prep_live_score.py. It will output a json file that could be directly used by the train or eval scripts. Unfortunately, I am not able to figure out what happens given the limited information provided. Have you decoded the mp4 streams to YUV? — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

[ccfco]

哥，有没有数据集给一份呢