Update conda env packages during the creation

triton-inference-server / dali_backend

The Triton backend that allows running GPU-accelerated data pre-processing pipelines implemented in DALI's python API.

MIT License

120 stars 28 forks source link

Closed JanuszL closed 1 year ago

JanuszL commented 1 year ago

when the conda environment is created it first uses the existing packages in the system - for example the one installed from Miniconda3-py38_4.12.0-Linux-x86_64.sh, which may become outdated and vulnerable sooner or later. This PR makes the environment to use the latest available necessary packages when created to fix all CVEs that are fixable by the update of the conda environment

Signed-off-by: Janusz Lisiecki jlisiecki@nvidia.com

dali-automaton commented 1 year ago

CI MESSAGE: [6923317]: BUILD STARTED

dali-automaton commented 1 year ago

CI MESSAGE: [6923317]: BUILD FAILED

dali-automaton commented 1 year ago

CI MESSAGE: [6923317]: BUILD PASSED