Closed njzjz closed 2 weeks ago
.github/workflows/test_cuda.yml (6)
`19-19`: The container configuration with NVIDIA CUDA image and GPU options is correctly set up for CUDA testing. --- `29-32`: Switching to `mamba-org/setup-micromamba` with a pinned version enhances the CI setup by efficiently managing Conda environments. --- `35-36`: Dynamically setting the TensorFlow and PyTorch root environment variables ensures correct paths are used in the workflow. --- `48-50`: Refactor the setting of `CMAKE_PREFIX_PATH` for clarity. --- `64-64`: The updated testing commands align well with the new environment setup and are correctly configured to use the `micromamba-shell`. --- `19-19`: Ensure that the 't4' runner is sufficient for the required tasks.Verification successful
--- The attempts to gather detailed specifications about the T4 GPU from NVIDIA's website have not yielded the required information. However, based on general knowledge, the NVIDIA T4 GPU is designed for high-performance computing and is CUDA-compatible, making it suitable for tasks requiring CUDA. Given this context, the 't4' runner should be sufficient for the required tasks in the GitHub Actions workflow.Scripts executed
The following scripts were executed for the analysis: --- Script: ```shell #!/bin/bash # Description: Verify the capabilities of the 't4' runner to handle the CUDA tasks. # Test: Check the specifications and performance benchmarks of the 't4' runner. echo "Check the 't4' runner specifications for CUDA compatibility and performance." ``` Length of output: 168 --- --- Script: ```shell #!/bin/bash # Description: Gather detailed specifications and capabilities of the 't4' runner. # Fetch the runner types supported by GitHub Actions. gh api -H "Accept: application/vnd.github.v3+json" /repos/:owner/:repo/actions/runners # Fetch the documentation or specifications for the 't4' runner. curl -s https://docs.github.com/en/actions/using-github-hosted-runners/about-github-hosted-runners | grep -i 't4' # Additionally, check NVIDIA's documentation for T4 specifications. curl -s https://www.nvidia.com/en-us/data-center/tesla-t4/ | grep -i 'specifications' ``` Length of output: 4931 --- --- Script: ```shell #!/bin/bash # Description: Gather detailed specifications and capabilities of the T4 GPU for CUDA tasks. # Fetch the NVIDIA T4 GPU specifications and check for CUDA compatibility. curl -s https://www.nvidia.com/en-us/data-center/tesla-t4/ | grep -A 10 -i 'NVIDIA T4 Specifications' ``` Length of output: 194 --- --- Script: ```shell #!/bin/bash # Description: Gather detailed specifications and capabilities of the T4 GPU for CUDA tasks. # Fetch the NVIDIA T4 GPU specifications and check for CUDA compatibility. curl -s https://www.nvidia.com/en-us/data-center/tesla-t4/ | grep -A 20 -i 'NVIDIA T4 Specifications' ``` Length of output: 313
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 82.71%. Comparing base (
d23f7ae
) to head (e555c83
). Report is 3 commits behind head on devel.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Transaction starting
critical libmamba Download error (28) Timeout was reached [https://conda.anaconda.org/conda-forge/linux-64/unzip-6.0-h7f98852_3.tar.bz2]
Operation too slow. Less than 30 bytes/sec transferred the last 60 seconds
Error: The process '/github/home/micromamba-bin/micromamba' failed with exit code 1
@caic99 any idea?
Transaction starting critical libmamba Download error (28) Timeout was reached [https://conda.anaconda.org/conda-forge/linux-64/unzip-6.0-h7f98852_3.tar.bz2] Operation too slow. Less than 30 bytes/sec transferred the last 60 seconds Error: The process '/github/home/micromamba-bin/micromamba' failed with exit code 1
@caic99 any idea?
Nothing but poor internet connection.
@njzjz Maybe you could use t4
runner for debugging purpose first.
@iProzd Where is this instance located?
Transaction starting critical libmamba Download error (28) Timeout was reached [https://conda.anaconda.org/conda-forge/linux-64/unzip-6.0-h7f98852_3.tar.bz2] Operation too slow. Less than 30 bytes/sec transferred the last 60 seconds Error: The process '/github/home/micromamba-bin/micromamba' failed with exit code 1
@caic99 any idea?
Nothing but poor internet connection.
@njzjz Maybe you could use
t4
runner for debugging purpose first.@iProzd Where is this instance located?
I see the same error with t4.
I see the same error with t4.
No ideas 😿
I decided to give up this PR and try another way.
Summary by CodeRabbit
New Features
Chores
mamba-org/setup-micromamba
for setting up the Python environment.