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 struct

GatedDeltaConv1dFwd

struct GatedDeltaConv1dFwd[write_state: Bool = True]

Gated DeltaNet causal conv1d forward pass (Pass 1 of two-pass prefill).

Reads/writes a single mutable conv-state pool of shape [max_slots, conv_dim, kernel_size-1] in place at slot slot_idx[batch_item]. No state-out tensor; the only output is the per-token conv output. The pool's dtype is independent of the working dtype, so the caller can keep the per-token tensors at fp32 while storing the pool at the model's native dtype (bf16).

Tensor Shapes: - conv_output_ragged : [total_seq_len, conv_dim] (OUT) - qkv_input_ragged : [total_seq_len, conv_dim] - conv_weight : [conv_dim, kernel_size] - conv_state : [max_slots, conv_dim, kernel_size-1] (MUT) - slot_idx : [batch_size] uint32 - input_row_offsets : [batch_size + 1] uint32

Parameters

• ​write_state (Bool): Whether to write the updated window back to conv_state. False only reads it for the look-back.

Implemented traits

AnyType, Deinitable, Movable

Methods

execute

static def execute[work_dtype: DType, state_dtype: DType, target: StringSpan[ImmStaticOrigin]](conv_output_ragged: ManagedTensorSlice[IOSpec[_, _].Output, static_spec=conv_output_ragged.static_spec], qkv_input_ragged: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=qkv_input_ragged.static_spec], conv_weight: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=conv_weight.static_spec], conv_state: ManagedTensorSlice[IOSpec[_, _].MutableInput, static_spec=conv_state.static_spec], slot_idx: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=slot_idx.static_spec], input_row_offsets: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=input_row_offsets.static_spec], ctx: DeviceContext)