Hi @vchung,
Can you please let me know why 9760782 failed?
Thanks a lot!
-Nikhil
Created by Nikhil Karthik nkck @nkck , ah that would make sense!
Let me defer your follow-up question over to @ktravaglini . Kyle, do you mind confirming whether the order of the sub-classes is the same as the training data? Thanks! Hi @vchung, Thank you it helps. The script runs fine locally, and very similar to my previous runs that were successful. I added a simple check that the subclasses are in the same order as in the training data. I think it must have returned "False" and hence didn't run the full script. Can you let me know (if fine to say) whether the ordered set of subclasses (obs/Subclass/categories) in the validation set is: ['Lamp5 Lhx6', 'Lamp5', 'Pax6', 'Sncg', 'Vip', 'Sst Chodl', 'Sst', 'Pvalb', 'Chandelier', 'L2/3 IT', 'L6 IT', 'L4 IT', 'L5 IT', 'L5 ET', 'L6 CT', 'L6b', 'L6 IT Car3', 'L5/6 NP', 'Astrocyte', 'OPC', 'Oligodendrocyte', 'Endothelial', 'VLMC', 'Microglia-PVM'] ? Sorry if there are much easier ways to deal with it via Anndata, but I'm not used to those and so using work arounds.
Thanks!
Nikhil Hi @nkck !
There doesn't appear to be any container logs for this submission. Can you confirm that you are able to run the container locally on your end and that it outputs the expected predictions file?
An example run you can use (that's similar to what's run on the submission system) is:
```
docker run \
--rm \
--network none \
--volume /PATH/TO/INPUT_FOLDER/:/input:ro \
--volume $PWD/output:/output:rw \
--memory 110G --shm-size 2G \
YOUR_DOCKER_IMAGE
```
Hope this helps!
EDIT: fix typo in docker command