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

nan_check_count

def nan_check_count[dtype: DType, rank: Int, target: StringSlice[StaticConstantOrigin]](nan_count_out: ManagedTensorSlice[Output, static_spec=nan_count_out.static_spec], inf_count_out: ManagedTensorSlice[Output, static_spec=inf_count_out.static_spec], input: ManagedTensorSlice[Input, static_spec=input.static_spec], ctx: DeviceContext)

Counts NaN/Inf values in a floating-point tensor.

Read-only: does not modify the input tensor. Outputs two single-element int32 tensors with the NaN and Inf counts. No D2H transfer or synchronization is performed — the counts stay on the same device as the input.

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