How to generate fireware about batch_size=n?

[zyt1024]

If I want to handle data with batch_size=2, can hls4ml generate the corresponding HLS code?

[vloncar]

No. You can manually modify the generated HLS to loop over the samples, but that may not be the most efficient implementation.

[manuelblancovalentin]

Opening this again, and just to clarify: during inference batch size is assumed to be always 1, right?

[vloncar]

Always 1