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
Currently, we only add tensors with different precision, e.g., Tensor_bf16 and Tensor_fp32. But we need to add NNTile dtypes into Python namespace at least for testing purposes, as relying on Numpy dtypes does not allow to run tests in mixed precision.
The text was updated successfully, but these errors were encountered:
Currently, we only add tensors with different precision, e.g.,
Tensor_bf16
andTensor_fp32
. But we need to add NNTile dtypes into Python namespace at least for testing purposes, as relying on Numpy dtypes does not allow to run tests in mixed precision.The text was updated successfully, but these errors were encountered: