Time too long for decoding images

[gcs790...@gmail.com]

Hi,

    I submitted my decoder about 11 hours ago. But I just get the following message:

    

    docker run -v "$(pwd)":"$(pwd)" -w "$(pwd)" clic2018/submit -n <team> -p <password> -e <email> <decoder> <images>

    Connecting...

    Sending data...

    Submission queued...

    Processing submission...

    Extracting decoder...

    Decoding images...

   Please check whether my decoder is still running on server.

    Thank you very much.

[gcs790...@gmail.com]

I can decode all images in 5 hours on my local machine. But it's too long to decode on server. Please check it for me, thank you very much. My team name is "test".

在 2018年4月12日星期四 UTC+8下午9:05:15，gcs790...@gmail.com写道：

[Lucas Theis]

This is the error that I'm seeing on the server:

./decode: line 2: 9 Killed python3 decoder.pyc

This is very likely because your decoder uses too much memory.

[Lucas Theis]

I am seeing the same error message for the team "test".

The amount of memory used can vary somewhat by machine, depending on what your decoder does (we've certainly observed this with tensorflow). If your decoder uses say 6GB memory locally, it may still cross the 8GB on the server.

On Apr 14, 2018, at 9:54 AM, gcs790...@gmail.com wrote:

Hi,

    I can run my decoder on my PC with 6G RAM, can you check it again for me?

在 2018年4月13日星期五 UTC+8下午5:18:44，ltheis写道：

This is the error that I'm seeing on the server:

./decode: line 2:     9 Killed                  python3 decoder.pyc

This is very likely because your decoder uses too much memory.

--
You received this message because you are subscribed to the Google Groups "Challenge on Learned Image Compression 2018" group.
To unsubscribe from this group and stop receiving emails from it, send an email to clic-2018+...@googlegroups.com.
To post to this group, send email to clic...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/clic-2018/e3c3b295-57c4-4fb1-a3cb-5fca5c6f02dd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[gcs790...@gmail.com]

Without running in the docker, our code just costs 6G RAM. But in docker ,it rise up to nearly 9GB. Can you increase the RAM for us? THX

在 2018年4月14日星期六 UTC+8下午5:08:23，ltheis写道：