build:release_gpu_linux --config=cuda_clang_official
test:release_gpu_linux --test_env=LD_LIBRARY_PATH="/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64"
# Local test jobs has to be 4 because parallel_gpu_execute is fragile, I think
test:release_gpu_linux --test_timeout=300,450,1200,3600 --local_test_jobs=4 --run_under=//tensorflow/tools/ci_build/gpu_build:parallel_gpu_execute

build:release_arm64_linux --config=release_linux_base

//
// Design decisions made:
// * Backing store for tensor memory has an associated deallocation
//   function.  This deallocation function will point to client code
//   for tensors populated by the client.  So the client can do things
//   like shadowing a numpy array.
// * We do not provide TF_OK since it is not strictly necessary and we
//   are not optimizing for convenience.
// * We make assumption that one session has one graph.  This should be