tabensemb.model.AbstractNN.bfloat16

method

AbstractNN.bfloat16() → T

Casts all floating point parameters and buffers to bfloat16 datatype.

Note

This method modifies the module in-place.

Returns:

self

Return type:

Module