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).

Python function

build_max_lengths_tensors

build_max_lengths_tensors()

max.nn.build_max_lengths_tensors(max_prompt_length, max_cache_length)

source

Builds two [1] uint32 scalar buffers of maximum lengths.

Parameters:

• max_prompt_length (int) – The maximum prompt (query) length.
• max_cache_length (int) – The maximum cache length.

Returns:

A tuple (max_prompt_length, max_cache_length) of Buffer, each of shape [1] and dtype uint32.

Return type:

tuple[Buffer, Buffer]