For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).
Mojo function
causal_conv1d_update_shape
def causal_conv1d_update_shape(input: T, conv_state_in: T, weight: T, bias: T) -> Tuple[IndexList[T.rank], IndexList[T.rank]]
Returns the output shapes for the causal_conv1d_update op.
The update produces two tensors: the convolution output for the new token(s) and the updated convolution state.
Args:
- βinput (
T): New input tokens with shape(batch, channels, seqlen). - βconv_state_in (
T): Previous convolution state with shape(batch, channels, state_len). - βweight (
T): Convolution weights with shape(channels, width). - βbias (
T): Per-channel bias with shape(channels,).
Returns:
Tuple[IndexList[T.rank], IndexList[T.rank]]: A tuple (output_shape, conv_state_shape) where output_shape
matches input.shape() and conv_state_shape matches
conv_state_in.shape().