[BUG] np.ndarray
of bfloat16 using ml_dtypes is being interpreted as complex64
#1075
Labels
np.ndarray
of bfloat16 using ml_dtypes is being interpreted as complex64
#1075
Describe the bug
np.ndarray
of bfloat16 using ml_dtypes is being interpreted as complex64 by mlx.To Reproduce
Expected behavior
Conversion to complex should not happen. Should remain as bfloat16
Desktop (please complete the following information):
Additional context
Originally posted by @lkarthee in #1066 (comment)
The text was updated successfully, but these errors were encountered: