[Feat] add chunk_gated_delta_rule triton support

python/sgl_kernel_npu/sgl_kernel_npu/fla/chunk.py

@@ -5,6 +5,18 @@
 
 import torch
 import torch.nn.functional as F
+from sgl_kernel_npu.fla.chunk_delta_h import (
+    chunk_gated_delta_rule_fwd_h_npu as chunk_gated_delta_rule_fwd_h,
+)
+from sgl_kernel_npu.fla.chunk_o import chunk_fwd_o_npu as chunk_fwd_o
+from sgl_kernel_npu.fla.chunk_scaled_dot_kkt import (
+    chunk_scaled_dot_kkt_fwd_npu as chunk_scaled_dot_kkt_fwd,
+)
+from sgl_kernel_npu.fla.cumsum import chunk_local_cumsum
+from sgl_kernel_npu.fla.l2norm import l2norm_fwd
+from sgl_kernel_npu.fla.solve_tril import solve_tril_npu as solve_tril
+from sgl_kernel_npu.fla.utils import SUPPRESS_LEVEL
+from sgl_kernel_npu.fla.wy_fast import recompute_w_u_fwd_npu as recompute_w_u_fwd
 
 
 def chunk_gated_delta_rule_native(
@@ -160,3 +172,181 @@ def chunk_gated_delta_rule_npu(
         final_cor_attn_out[:, start:end, ...] = core_attn_out[b_idx]
 
     return final_cor_attn_out, last_recurrent_state
+
+
+def chunk_gated_delta_rule_fwd(
+    q: torch.Tensor,
+    k: torch.Tensor,
+    v: torch.Tensor,
+    g: torch.Tensor,
+    beta: torch.Tensor,
+    scale: float,
+    initial_state: torch.Tensor,
+    output_final_state: bool,
+    cu_seqlens: Optional[torch.LongTensor] = None,
+):
+    g = chunk_local_cumsum(g, chunk_size=64, cu_seqlens=cu_seqlens)
+    # obtain WY representation. u is actually the new v.
+    A = chunk_scaled_dot_kkt_fwd(
+        k=k, beta=beta, g_cumsum=g, cu_seqlens=cu_seqlens, output_dtype=torch.float32
+    )
+    A = solve_tril(A=A, cu_seqlens=cu_seqlens, output_dtype=k.dtype)
+    w, u = recompute_w_u_fwd(
+        k=k,
+        v=v,
+        beta=beta,
+        A=A,
+        g_cumsum=g,
+        cu_seqlens=cu_seqlens,
+    )
+    h, v_new, final_state = chunk_gated_delta_rule_fwd_h(
+        k=k,
+        w=w,
+        u=u,
+        g=g,
+        initial_state=initial_state,
+        output_final_state=output_final_state,
+        cu_seqlens=cu_seqlens,
+    )
+    o = chunk_fwd_o(
+        q=q,
+        k=k,
+        v=v_new,
+        h=h,
+        g=g,
+        scale=scale,
+        cu_seqlens=cu_seqlens,
+    )
+    if SUPPRESS_LEVEL < 3:
+        return g, o, A, final_state, None, None, None
+    elif SUPPRESS_LEVEL >= 3:
+        return g, o, A, final_state, w, h, v_new
+
+
+@torch.compiler.disable
+def chunk_gated_delta_rule(
+    q: torch.Tensor,
+    k: torch.Tensor,
+    v: torch.Tensor,
+    g: torch.Tensor,
+    beta: torch.Tensor,
+    scale: float = None,
+    initial_state: torch.Tensor = None,
+    output_final_state: bool = False,
+    cu_seqlens: Optional[torch.LongTensor] = None,
+    head_first: bool = False,
+    use_qk_l2norm_in_kernel: bool = False,
+):
+    r"""
+    Args:
+        q (torch.Tensor):
+            queries of shape `[B, T, H, K]` if `head_first=False` else `[B, H, T, K]`.
+        k (torch.Tensor):
+            keys of shape `[B, T, H, K]` if `head_first=False` else `[B, H, T, K]`.
+        v (torch.Tensor):
+            values of shape `[B, T, H, V]` if `head_first=False` else `[B, H, T, V]`.
+        g (torch.Tensor):
+            (forget) gating tensor (in log space!) of shape `[B, T, H]` if `head_first=False` else `[B, H, T]`.
+        beta (torch.Tensor):
+            betas of shape `[B, T, H]` if `head_first=False` else `[B, H, T]`.
+        scale (Optional[int]):
+            Scale factor for the RetNet attention scores.
+            If not provided, it will default to `1 / sqrt(K)`. Default: `None`.
+        initial_state (Optional[torch.Tensor]):
+            Initial state of shape `[N, H, K, V]` for `N` input sequences.
+            For equal-length input sequences, `N` equals the batch size `B`.
+            Default: `None`.
+        output_final_state (Optional[bool]):
+            Whether to output the final state of shape `[N, H, K, V]`. Default: `False`.
+        cu_seqlens (torch.LongTensor):
+            Cumulative sequence lengths of shape `[N+1]` used for variable-length training,
+            consistent with the FlashAttention API.
+        head_first (Optional[bool]):
+            Whether the inputs are in the head-first format, which is not supported for variable-length inputs.
+            Default: `False`.
+
+    Returns:
+        o (torch.Tensor):
+            Outputs of shape `[B, T, H, V]` if `head_first=False` else `[B, H, T, V]`.
+        final_state (torch.Tensor):
+            Final state of shape `[N, H, K, V]` if `output_final_state=True` else `None`.
+
+    Examples::
+        >>> import torch
+        >>> import torch.nn.functional as F
+        >>> from einops import rearrange
+        >>> from fla.ops.gated_delta_rule import chunk_gated_delta_rule
+        # inputs with equal lengths
+        >>> B, T, H, K, V = 4, 2048, 4, 512, 512
+        >>> q = torch.randn(B, T, H, K, dtype=torch.bfloat16, device='cuda')
+        >>> k = F.normalize(torch.randn(B, T, H, K, dtype=torch.bfloat16, device='cuda'), p=2, dim=-1)
+        >>> v = torch.randn(B, T, H, V, dtype=torch.bfloat16, device='cuda')
+        >>> beta = torch.rand(B, T, H, dtype=torch.bfloat16, device='cuda').sigmoid()
+        >>> g = F.logsigmoid(torch.rand(B, T, H, dtype=torch.bfloat16, device='cuda'))
+        >>> h0 = torch.randn(B, H, K, V, dtype=torch.bfloat16, device='cuda')
+        >>> o, ht = chunk_gated_delta_rule(
+            q, k, v, g, beta,
+            initial_state=h0,
+            output_final_state=True
+        )
+        # for variable-length inputs, the batch size `B` is expected to be 1 and `cu_seqlens` is required
+        >>> q, k, v, beta, g = map(lambda x: rearrange(x, 'b t ... -> 1 (b t) ...'), (q, k, v, beta, g))
+        # for a batch with 4 sequences, `cu_seqlens` with 5 start/end positions are expected
+        >>> cu_seqlens = q.new_tensor([0, 2048, 4096, 6144, 8192], dtype=torch.long)
+        >>> o_var, ht_var = chunk_gated_delta_rule(
+            q, k, v, g, beta,
+            initial_state=h0,
+            output_final_state=True,
+            cu_seqlens=cu_seqlens
+        )
+    """
+    from einops import rearrange
+
+    assert q.dtype == k.dtype == v.dtype
+    assert (
+        q.dtype != torch.float32
+    ), "ChunkGatedDeltaRuleFunction does not support float32. Please use bfloat16."
+    assert (
+        len(beta.shape) == 3
+    ), "beta must be of shape [B, T, H] if head_first=False, or [B, H, T] otherwise."
+
+    if head_first:
+        raise DeprecationWarning(
+            "head_first is deprecated and will be removed in a future version. "
+            "Please use head_first=False for now instead."
+        )
+        q, k, v, beta, g = map(
+            lambda x: rearrange(x, "b h t ... -> b t h ..."), (q, k, v, beta, g)
+        )
+    # if not head_first and q.shape[1] < q.shape[2]:
+    #     warnings.warn(
+    #         f"Input tensor shape suggests potential format mismatch: seq_len ({q.shape[1]}) < num_heads ({q.shape[2]}). "
+    #         "This may indicate the inputs were passed in head-first format [B, H, T, ...] "
+    #         "when head_first=False was specified. "
+    #         "Please verify your input tensor format matches the expected shape [B, T, H, ...]."
+    #     )
+    if cu_seqlens is not None:
+        if q.shape[0] != 1:
+            raise ValueError(
+                f"The batch size is expected to be 1 rather than {q.shape[0]} when using `cu_seqlens`."
+                f"Please flatten variable-length inputs before processing."
+            )
+        if initial_state is not None and initial_state.shape[0] != len(cu_seqlens) - 1:
+            raise ValueError(
+                f"The number of initial states is expected to be equal to the number of input sequences, "
+                f"i.e., {len(cu_seqlens) - 1} rather than {initial_state.shape[0]}."
+            )
+    if scale is None:
+        scale = k.shape[-1] ** -0.5
+
+    if use_qk_l2norm_in_kernel:
+        q = l2norm_fwd(q)
+        k = l2norm_fwd(k)
+
+    _, o, _, final_state, _, _, _ = chunk_gated_delta_rule_fwd(
+        q, k, v, g, beta, scale, initial_state, output_final_state, cu_seqlens
+    )
+    o = o.to(q.dtype)
+    if head_first:
+        o = rearrange(o, "b t h ... -> b h t ...")
+    return o, final_state