-
Notifications
You must be signed in to change notification settings - Fork 916
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[BUG] cudaErrorIllegalAddress
when trying to write a large list column to json
#13030
Comments
Thank you @galipremsagar for raising this issue. Using the 23.02 release image, I reproduced your issue with List types:
I also identified a potentially related error with a simpler list column:
and also...
|
I believe that strings column and lists column are processed differently. So you may still have OOM with lists while not with strings.
|
Thanks @ttnghia import cudf
gdf = cudf.DataFrame({"lists": [[1, 2, 3], [4]]*5})
gdf.to_json("cudf.json", engine="cudf", orient='records', lines=True, rows_per_chunk=8) |
Fixes the OOM access error while using chunked JSON writer on list columns. The issue is present in struct columns also, which is fixed in this change. Fixes #13030 Authors: - Karthikeyan (https://github.com/karthikeyann) - GALI PREM SAGAR (https://github.com/galipremsagar) Approvers: - Nghia Truong (https://github.com/ttnghia) - David Wendt (https://github.com/davidwendt) URL: #13108
Describe the bug
When we try writing a dataframe that has
ListColumn
with large number of rows, we are getting a runtime error that sayscudaErrorIllegalAddress
.Steps/Code to reproduce bug
with small number of rows it works fine:
Right as we increase the number of rows we see the issue:
Somethings to note:
Expected behavior
Should be able to write the column to dataframe.
Environment overview (please complete the following information)
Environment details
Please run and paste the output of the
cudf/print_env.sh
script here, to gather any other relevant environment detailsClick here to see environment details
Additional context
Add any other context about the problem here.
The text was updated successfully, but these errors were encountered: