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 struct
GatedDeltaRecurrenceFwd
struct GatedDeltaRecurrenceFwd
Gated DeltaNet recurrence forward pass (Pass 2 of two-pass prefill).
Reads/writes a single mutable recurrent-state pool of shape
[max_slots, num_value_heads, key_head_dim, value_head_dim] in place
at slot slot_idx[batch_item]. No state-out tensor; the only output
is the per-token recurrence output.
Tensor Shapes: - recurrence_output : [total_seq_len, value_dim] (OUT) - qkv_conv_output : [total_seq_len, conv_dim] - decay_per_token : [total_seq_len, num_value_heads] - beta_per_token : [total_seq_len, num_value_heads] - recurrent_state : [max_slots, num_value_heads, KD, VD] (MUT) - slot_idx : [batch_size] uint32 - input_row_offsets : [batch_size + 1] uint32
Implemented traits
Methods
execute
static def execute[work_dtype: DType, state_dtype: DType, target: StringSlice[ImmStaticOrigin]](recurrence_output: ManagedTensorSlice[IOSpec[_, _].Output, static_spec=recurrence_output.static_spec], qkv_conv_output: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=qkv_conv_output.static_spec], decay_per_token: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=decay_per_token.static_spec], beta_per_token: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=beta_per_token.static_spec], recurrent_state: ManagedTensorSlice[IOSpec[_, _].MutableInput, static_spec=recurrent_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)
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!