Python function

float32_array_to_buffer

`float32_array_to_buffer()`

max.pipelines.lib.float32_array_to_buffer(array, *, dtype, device)

source

Create a device buffer from float32 host data without a cast-only graph.

For bfloat16, the host array is converted to bf16 bit-patterns stored as uint16 and then reinterpreted as bfloat16 in MAX.

Parameters:

Return type:

Buffer