Why minimum Byte Size Forced to 2 when calculation hidden bytes for FP8 in get_hidden_bytes()

in https://github.com/deepseek-ai/DeepEP/blob/main/README.md Interfaces and examples section,

def get_hidden_bytes(x: torch.Tensor) -> int:
t = x[0] if isinstance(x, tuple) else x
return t.size(1) * max(t.element_size(), 2)

FP8 tensors use 1 byte per element, but the function forces a 2-byte minimum，Is there a memory/performance trade-off?

Read README.md's Interfaces and examples section and inspect get_hidden_bytes(). Start by tracing how its result is used for FP8 tensors and compare the memory and performance implications of the two-byte minimum. Done means the rationale for max(t.element_size(), 2) is established and clearly explained in the issue or documentation.

Written by the indexing model from the issue text.