search

rapidsai / clx

A collection of RAPIDS examples for security analysts, data scientists, and engineers to quickly get started applying RAPIDS and GPU acceleration to real-world cybersecurity use cases.
Apache License 2.0
168 stars 68 forks source link

Update `conda` environment name for CI #436

Closed ajschmidt8 closed 3 years ago

ajschmidt8 commented 3 years ago

The gdf conda environment has been replaced with the rapids environment. A symlink was put in place for gdf to continue to work, but the symlink will be removed in the near future. This PR updates all scripts to use the rapids environment name.

ajschmidt8 commented 3 years ago

@efajardo-nv and/or @BartleyR, can you review/test the changes in https://github.com/rapidsai/clx/pull/436/commits/a2250926645b86cf9b7baba840f1f15d70b7d20c? I believe that was causing the CI issues seen in my previous commit (log link). CI is running with Python 3.8, but gwerbin only has a 3.7 version of seqeval on Anaconda. Therefore conda was installing the seqeval 0.0.12 version from conda-forge, which was downgrading arrow to 1.0.1, which we no longer use as of https://github.com/rapidsai/cudf/pull/7495. From my local testing, upgrading to seqeval 1.2.2 seems to prevent arrow from being downgraded.

Here's the changelog related to seqeval 1.2.2 for reference: https://github.com/chakki-works/seqeval/releases/tag/v1.2.2

ajschmidt8 commented 3 years ago

Actually, upon further inspection, seqeval 1.2.2 is what's in the Docker images by default right now (see screenshot below). And the cybert_example_training.ipynb notebook seems to run fine there. So I think this change should be fine.

image

ajschmidt8 commented 3 years ago

@gpucibot merge