# Pass smoke test but fail normal submission **URL:** \ **Category:** SNOMED CT Entity Linking\ **Created:** [February 17, 2024, 1:23am UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325 "2024-02-17T01:23:14Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![hl2659](https://avatars.discourse-cdn.com/v4/letter/h/9e8a1a/32.png) [@hl2659](https://community.drivendata.org/u/hl2659)\ **Post date:** [February 17, 2024, 1:23am UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/1 "2024-02-17T01:23:14Z")
My submission pass the smoke test but fail with the normal submission. And the log for normal submission stopped at “python main.py” making it impossible to debug. Can anyone help? ---
**Author:** ![chrisk-dd](https://avatars.discourse-cdn.com/v4/letter/c/b5ac83/32.png) [@chrisk-dd](https://community.drivendata.org/u/chrisk-dd)\ **Post date:** [February 17, 2024, 4:23pm UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/2 "2024-02-17T16:23:25Z")
Hey @hl2659 - I will follow up with you via DM to share logs of the error with you. ---
**Author:** ![DHC\_chinese](https://avatars.discourse-cdn.com/v4/letter/d/898d66/32.png) [@DHC\_chinese](https://community.drivendata.org/u/DHC_chinese)\ **Post date:** [February 26, 2024, 4:27pm UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/3 "2024-02-26T16:27:10Z")
The same problem also happened to me ---
**Author:** ![chrisk-dd](https://avatars.discourse-cdn.com/v4/letter/c/b5ac83/32.png) [@chrisk-dd](https://community.drivendata.org/u/chrisk-dd)\ **Post date:** [February 26, 2024, 6:04pm UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/4 "2024-02-26T18:04:17Z")
@DHC_chinese I will follow up with you via DM. For others who may have a similar issue, it may be related to the fact that the smoke test dataset is relatively small, and that your code encounters edge cases on the larger test set that aren’t present in the smoke test dataset. I would recommend that if you are experiencing similar issues (smoke tests passing but full submissions failing) that you consider expanding the set of smoke test notes you are using locally to encompass more of the training set notes. ---
**Author:** ![snakesgun](https://avatars.discourse-cdn.com/v4/letter/s/6de8d8/32.png) [@snakesgun](https://community.drivendata.org/u/snakesgun)\ **Post date:** [March 1, 2024, 9:56am UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/5 "2024-03-01T09:56:49Z")
Same issue happened to me. ---
**Author:** ![fganong](https://avatars.discourse-cdn.com/v4/letter/f/96bed5/32.png) [@fganong](https://community.drivendata.org/u/fganong)\ **Post date:** [March 10, 2026, 2:06pm UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/7 "2026-03-10T14:06:45Z")
is there a convenient way to run the submission.zip locally on all the test data? (ie. to check for the edge cases mentioned above). I can edit my code to point at different data; but I’d prefer to run the exact submission.zip. Thanks for any help! ---
**Author:** ![fganong](https://avatars.discourse-cdn.com/v4/letter/f/96bed5/32.png) [@fganong](https://community.drivendata.org/u/fganong)\ **Post date:** [March 10, 2026, 3:11pm UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/8 "2026-03-10T15:11:57Z")
oops, I now see this is a dumb question, of course I can’t run locally on the hidden data. I can run on the training data; so I need to modify my script (slightly) to find the downloaded training data. ---
**Author:** ![chrisk-dd](https://avatars.discourse-cdn.com/v4/letter/c/b5ac83/32.png) [@chrisk-dd](https://community.drivendata.org/u/chrisk-dd)\ **Post date:** [March 16, 2026, 2:07pm UTC](https://community.drivendata.org/t/pass-smoke-test-but-fail-normal-submission/10325/10 "2026-03-16T14:07:24Z")
@fganong no such thing as a dumb question, thank you for learning in public (and by doing so helping others)!