Skip to content
Test CUDA Builds

Import torch only when the Python runtime needs it, warn before outputs follow inputs, and upload bfloat16 without torch #22239

Sign in to view logs

Import torch only when the Python runtime needs it, warn before outputs follow inputs, and upload bfloat16 without torch

Import torch only when the Python runtime needs it, warn before outputs follow inputs, and upload bfloat16 without torch #22239

Run time

Learn about OS pricing on GitHub Actions
Job Run time
Validate benchmark dispatch inputs
2s
Get changed files / get-changed-files
7s
CI run decision / decide
36s
test-executorch-cuda-build-13.0 / linux-job
22m 0s
unittest-cuda (kernel) / linux-job
1h 9m 17s
test-executorch-cuda-build-13.4 / linux-job
24m 48s
unittest-cuda (system) / linux-job
1h 6m 27s
unittest-cuda-runtime
-1s
test-model-cuda-e2e-${{ matrix.model.name }}-${{ matrix.quant }}
-1s
test-models-cuda
-1s
export-cuda-multi-arch-a100
-1s
upload-benchmark-results
-1s
test-muse-glimmer-cuda-e2e-${{ matrix.variant }}-${{ matrix.mode }}
-1s
test-cuda-merged-pte-a100
-1s
merge-and-test-cuda-multi-arch-a10g
-1s
unittest-cuda-quantized-gemm (${{ matrix.shard }})
-1s
unittest-cuda-muse-glimmer-batching
-1s
check-all-cuda-builds
4s
3h 3m 11s