torch.rand can sample the upper bound for lower precision floating point dtypes on CUDA #96947
Labels
high priority
module: bfloat16
module: correctness (silent)
issue that returns an incorrect result silently
module: half
Related to float16 half-precision floats
module: random
Related to random number generation in PyTorch (rng generator)
triaged
This issue has been looked at a team member, and triaged and prioritized into an appropriate module
Documentation:
Failures happen for
float16
,bfloat16
, andcomplex32
only on CUDA.This was detected in #96331, which uses
Tensor.uniform_
under the hood, but I guess internally it is the same kernel.cc @ezyang @gchanan @zou3519 @pbelevich
The text was updated successfully, but these errors were encountered: