For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).
Mojo function
mhc_post_block_size
def mhc_post_block_size[dtype: DType, hidden_size: Int]() -> Int
Returns the threads per block mhc_post takes for hidden_size.
256 where their 16-byte vectors split the hidden axis into whole blocks, else 128.
Parameters:
Returns:
Int: The threads per block.