Dear PyTorch users,
We are still working on a version of the starting kit including PyTorch example and we apologize that it has not been finalized yet.
Meanwhile, if you want to try out submissions in PyTorch, you are much welcome to refer to Kakaobrain's code (they are 1st winner of AutoCV1 and they use PyTorch together with some Tensorflow). Their github repo is here: https://github.com/kakaobrain/autoclint.
Best regards,
The organizing team
Hi Committee,
Thanks for organizing this awesome challenge. I’m trying PyTorch code submission. It could work fine when I did the local test, but failed on the server without providing any useful information. Would you mind to provide some suggestions? My submission ID is 1823.
In addition, is it possible to extend the competition deadline, because it could take more time to handle and be familiar to TF-data for PyTorch users compared to that of TF users?
Dear participant,
Your code encountered following error:
Traceback (most recent call last):
File "/tmp/codalab/tmpSmJTCl/run/ingestion_program/ingestion.py", line 293, in <module>
from model import Model # in participants' model.py
File "/tmp/codalab/tmpSmJTCl/run/submission/model.py", line 11, in <module>
from architectures import get_model
ImportError: cannot import name 'get_model'
This error message can be found in the Output Log of Ingestion Step, after clicking on the button "+" of your corresponding submission in "My Submissions" tab.
Best,
The AutoDL challenges organizers
Thanks for your help. I did not find the python error information in both Output log and Error log. Did I do something incorrectly? Below are the information shown in my Output log and Error log.
I tried to open the Output log (the link should be https://autodl.lri.fr/my/competition/submission/1823/stdout.txt), but it only shows
/tmp/codalab/tmpEs_PyK/run/program
/tmp/codalab/tmpEs_PyK/run/program/default_scores.txt
/tmp/codalab/tmpEs_PyK/run/program/default_curve.png
2019-08-03 20:03:31,254 INFO evaluate.py: ##################################################
2019-08-03 20:03:31,254 INFO evaluate.py: Version: v20190505
2019-08-03 20:03:31,254 INFO evaluate.py: This is the parent scoring program. It reads from input_dir/res_i/ all partial results from children phases, and outputs aggregated learning curves and scores to output_dir.
2019-08-03 20:03:31,254 INFO evaluate.py: ##################################################
Copying input folder....
2019-08-03 20:03:31,260 INFO evaluate.py: Checking /tmp/codalab/tmpEs_PyK/run/input/res_2
2019-08-03 20:03:31,261 INFO evaluate.py: Checking /tmp/codalab/tmpEs_PyK/run/input/res_3
2019-08-03 20:03:31,261 INFO evaluate.py: Checking /tmp/codalab/tmpEs_PyK/run/input/res_4
2019-08-03 20:03:31,261 INFO evaluate.py: Checking /tmp/codalab/tmpEs_PyK/run/input/res_5
2019-08-03 20:03:31,262 INFO evaluate.py: Checking /tmp/codalab/tmpEs_PyK/run/input/res_6
2019-08-03 20:03:31,262 INFO evaluate.py: [+] Results validation done.
2019-08-03 20:03:31,265 INFO evaluate.py: [+] Score reading done.
2019-08-03 20:03:31,265 INFO evaluate.py: Score list: [-0.999999, -0.999999, -0.999999, -0.999999, -0.999999]
2019-08-03 20:03:31,265 INFO evaluate.py: [+] Score writing done.
2019-08-03 20:03:31,265 INFO evaluate.py: [+] Learning curve reading done.
2019-08-03 20:03:31,266 INFO evaluate.py: [+] Learning curve writing done.
2019-08-03 20:03:31,266 INFO evaluate.py: [+] Parent scoring program finished!
I also tried to open the Error Log, it (https://autodl.lri.fr/my/competition/submission/1823/stderr.txt) shows:
WARNING: Your kernel does not support swap limit capabilities or the cgroup is not mounted. Memory limited without swap.
WARNING:root:/tmp/codalab/tmpEs_PyK/run/input/res_2 does not exist. Default values will be used.
WARNING:root:/tmp/codalab/tmpEs_PyK/run/input/res_3 does not exist. Default values will be used.
WARNING:root:/tmp/codalab/tmpEs_PyK/run/input/res_4 does not exist. Default values will be used.
WARNING:root:/tmp/codalab/tmpEs_PyK/run/input/res_5 does not exist. Default values will be used.
WARNING:root:/tmp/codalab/tmpEs_PyK/run/input/res_6 does not exist. Default values will be used.
Oh, I missed the "Ingestion Step", I did not find the "Output Log" in my "Ingestion Step". Actually, my "Ingestion Step" and "Prediction Step" are empty, only "Scoring Step" has the log bottom. Is there anything wrong?
Posted by: DXY0808 @ Aug. 7, 2019, 5:25 p.m.Sorry to bother you. I found the Output log of “Ingestion Step” in the results of each individual dataset. Many thanks for your help.
Posted by: DXY0808 @ Aug. 8, 2019, 7:35 a.m.Yes I forgot to mention that you need to click on say, "Dataset 1"
Posted by: Zhengying @ Aug. 8, 2019, 8:04 a.m.