Hi, We have experienced a weird behavior on one of the networks that we were learning and were expecting to obtain the models from today. The logs of the submission with ID 7874358 of docker syn7827568 show that the docker was working correctly where today, out of a sudden, all the models were lost and the learning have restarted from the beginning (after 4 days of waiting to obtain the models). In fact if you revise the log files from the history, you will see the change happens between log version 269 and 270. Please note that the docker was on the final fold of the learning and all stored models are unavailable and the learning started from scratch. it seems like wither the docker or server was reset without notifying us. Such a problem is now forcing us to drop one of the submissions we planned to submit to the scoring. 1- Can we get more details about the problem that occurred or why was our docker reset? 2- did this effect other users in the challenge ? because it seems like a server problem

Created by wissam Baddar wbaddar
> 1- Can we get more details about the problem that occurred or why was our docker reset? Two days ago the server that was running your submission (7874358) experienced a problem and the state of the container was lost. In such a case we restart the submission on another server. At this time we do not know conclusively why the problem occurred, it's still under investigation. > 2- did this effect other users in the challenge ? because it seems like a server problem Though it is not common, is not the sole server problem we have experienced. We follow a consistent policy of restarting submissions when a server problem is detected.   We apologize for the inconvenience.

Restarting a docker on the training lane page is loading…