I have the following structure:
```
/usr/local/bin :
| run_model.py
| model.pkl
|
| -- --- names..csv
| --
Created by Lorenzo de Trizio Lorenzo.deTrizio Hello @Lorenzo.deTrizio ,
Regarding your question - https://www.synapse.org/Synapse:syn52817032/discussion/threadId=11437&replyId=33051
If you are asking about data availability after the challenge has concluded, then yes, the data can be made available upon request. You will need to submit a request to collaborate with the PEGS Study as specified on this webpage - https://www.niehs.nih.gov/research/atniehs/labs/crb/studies/pegs/collaboration/proposal
@vchung
thanks for the reply! @Lorenzo.deTrizio ,
Your Docker submissions are run on the real data, not synthetic, so in this case, we expect there to be more than 697 sample/participant IDs for the Leaderboard Round.
Hope this helps! @vchung
regarding the error message: the validation file has 697 participant ID, the message show that 2365 missing ID: I wrong somethings? @vchung
i have a request: is possible to posticipate the data for the end for leaderboard submission ?
Will be great if the answer was Yes :) @vchung Thanks for the your precise support! :) @Lorenzo.deTrizio ,
Your model is expected to generate a prediction for every sample ID or participant ID for PEGS participants. The validation error you received indicates that a prediction was not found those sample/participant IDs. hello @vchung thanks.
Now I have this:
Found 2365 missing ID(s): ['10012388', '10040779', '10060133', '10091736', '10114784', '10121994', '10134148', '10134153', '10148239', '10150483', '10166321', '10169035', '10176906', '10186687', '10208711', '10218034', '10230153', '10247523', '10263319', '10268423', '10270606', '10272037', '10278178', '10278208', '10307929', '10312601', '10324894', '10336049', '10347325', '10360095', '10362308', '10367479', '10374544', '10378666', '10380801', '10384278', '10387326', '10394263', '10395126', '...'.
What does it mean? @Lorenzo.deTrizio ,
Right, so the Leaderboard is showing the "Expected 1 Docker container output file with base name 'predictions' in the output directory. Got 0..." error message because a predictions file could not be found. And the predictions file could not be found because your Docker submission was not able to run, due to the aforementioned error.
> How I can get the right error log?
If you would like to see the logs of your Docker runs, please go to the [leaderboard table](syn57373526) and scroll to the right until you see the **docker_logs_id** column. Provided in this column are links to the log files. If you click on one of the links, you will be taken to a page where you can download the log file for a particular submission.
> I use a Mac Book pro M3 pro: is necessary to create the docker for AMD64?
Yes, our submission system is run on a AWS EC2, which uses the x86-64 cpu architecture, so it will be necessary to build the image for x86. Thanks again, but in right part of the leaderboard have another error.
How I can get the right error log?
I use a Mac Book pro M3 pro: is necessary to create the docker for AMD64? Hi @Lorenzo.deTrizio,
I checked the Docker logs for submission ID 9748049 and it shows the following error again:
```
exec /usr/local/bin/python: exec format error
``` Hello @vchung ,
thanks for the support: I have modified the script but I always have this: Evaluation failed for Submission 9748049.
Reason: 'Expected 1 Docker container output file with base name 'predictions' in the output directory. Got 0, or a file incorrectly named. If multiple output files are generated, please zip them into a single file for processing'.
Pls I can send the docker file, the script if you can support me: have a email where I can send the file?
I need to post a valid submission,
Lorenzo @Lorenzo.deTrizio ,
We will be mounting `/output` when running your Docker container, so it will already be available - no need to create it. Thanks for reply @vchung : the forlder exist or is necessary to create it? Hi, @Lorenzo.deTrizio:
Thank you for sharing your file structure! As noted in the Submission Tutorial, the final output should be saved to `/output`, that is: root folder -> output. From your shared file tree, it looks like `predictions.csv` was instead saved to `/usr/local/bin/output` instead. This would explain the error you received about not being able to find the predictions file in the output folder.
Hope this helps!