For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo function

annotate_ops

annotate_ops(ops: List[OpDesc], model: TargetCostModel) -> List[OpDesc]

Apply a target cost model to logical ops.

For each op with a kernel-specific tag (< 128), looks up the cost model and fills in resource, latency, and role. Infrastructure ops (tag >= 128) are passed through unchanged — they carry their own annotations.

This is the bridge between the algorithm (which declares logical ops via OpDesc.logical()) and the scheduler (which needs resource/latency data).

Returns:

List[OpDesc]

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit