load_conv1d_kernels

load_conv1d_kernels()

Import the 1D non-causal kernels now instead of on first call.

Same purpose as load_fused_fft_conv2d(): a torch.compile trace that runs before any eager call must find the operators already registered and the explicit_conv1d dispatch table already resolved.

Raises:

ImportError – If subquadratic_ops_torch is not installed.

Return type:

None