flashinfer.cake_causal_conv1d_nondet_train.cake_causal_conv1d_nondet_train_forward_workspace

flashinfer.cake_causal_conv1d_nondet_train.cake_causal_conv1d_nondet_train_forward_workspace(S: int, C: int, device: device) → Tensor

Allocate a zeroed forward workspace tensor.

Parameters:
  • S (int) – Packed token count.

  • C (int) – Channel count.

  • device (torch.device) – CUDA device that will execute the kernel.

Returns:

Zeroed int32 forward workspace for reuse on a single stream.

Return type:

torch.Tensor