You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
At the moment we're using in a couple of placesCUDA.functional() to check whether CUDA is available. However that refers to CUDA.jl only, using hermetic CUDA builds of XLA (🥲) in Reactant means that that check may disagree with what's actually available in Reactant (notable current cases: CUDA < 12.1 on x86_64-linux, or any CUDA GPU on aarch64-linux). This causes tests to fail on such systems, because for example at
At the moment we're using in a couple of places
CUDA.functional()
to check whether CUDA is available. However that refers toCUDA.jl
only, using hermetic CUDA builds of XLA (🥲) in Reactant means that that check may disagree with what's actually available in Reactant (notable current cases: CUDA < 12.1 on x86_64-linux, or any CUDA GPU on aarch64-linux). This causes tests to fail on such systems, because for example atReactant.jl/test/integration/cuda.jl
Line 30 in 29627db
@mofeing suggested something like
The text was updated successfully, but these errors were encountered: