IMPORTANT: To view this page as Markdown, append `.md` to the URL (e.g. /max/get-started.md). For the complete documentation index, see llms.txt.
Skip to main content
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

causal_conv1d_varlen_update_cpu

causal_conv1d_varlen_update_cpu[x_dtype: DType, weight_dtype: DType, bias_dtype: DType, output_dtype: DType, conv_state_dtype: DType, cache_seqlens_dtype: DType, conv_state_indices_dtype: DType](batch: Int, dim: Int, seqlen: Int, width: Int, state_len: Int, x: TileTensor[x_dtype, address_space=x.address_space, linear_idx_type=x.linear_idx_type, element_size=x.element_size], weight: TileTensor[weight_dtype, address_space=weight.address_space, linear_idx_type=weight.linear_idx_type, element_size=weight.element_size], bias: TileTensor[bias_dtype, address_space=bias.address_space, linear_idx_type=bias.linear_idx_type, element_size=bias.element_size], conv_state: TileTensor[conv_state_dtype, address_space=conv_state.address_space, linear_idx_type=conv_state.linear_idx_type, element_size=conv_state.element_size], cache_seqlens: TileTensor[cache_seqlens_dtype, address_space=cache_seqlens.address_space, linear_idx_type=cache_seqlens.linear_idx_type, element_size=cache_seqlens.element_size], conv_state_indices: TileTensor[conv_state_indices_dtype, address_space=conv_state_indices.address_space, linear_idx_type=conv_state_indices.linear_idx_type, element_size=conv_state_indices.element_size], output: TileTensor[output_dtype, address_space=output.address_space, linear_idx_type=output.linear_idx_type, element_size=output.element_size], x_batch_stride: UInt32, x_dim_stride: UInt32, x_seqlen_stride: UInt32, weight_dim_stride: UInt32, weight_width_stride: UInt32, conv_state_batch_stride: UInt32, conv_state_dim_stride: UInt32, conv_state_seqlen_stride: UInt32, out_batch_stride: UInt32, out_dim_stride: UInt32, out_seqlen_stride: UInt32, silu_activation: Bool, pad_slot_id: Int32, has_conv_state_indices: Bool, has_cache_seqlens: Bool, has_bias: Bool)

Update function for causal conv1d decode.

Updates the convolution state and computes output for decode steps. Supports circular buffer state management with cache_seqlens.