Closed mattdangerw closed 11 months ago
Patch coverage: 100.00%
and no project coverage change.
Comparison is base (
e8db3b6
) 76.56% compared to head (b4358ca
) 76.56%.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Making multi backend dtype comparisons for code based on keras-core is tricky. Torch has dtype classes such that
torch_tensor.dtype is string_dtype
will always be false.We use
standardize_dtype
extensively in our own code and should expose it to clients.is_float_dtype
andis_int_dtype
we can include as useful helpers.