Support GLM-5.2 (glm_moe_dsa, DeepSeek-V3.2-style DSA MoE)

docs/CN/source/models/supported_models.rst

@@ -56,6 +56,8 @@ lightllm 支持大多数的主流的开源大语言模型以及多模态模型
     - 
   * - `Qwen3-Moe <https://github.com/QwenLM/Qwen3>`_ 
     - 
+  * - `GLM-5.2 <https://huggingface.co/zai-org/GLM-5.2>`_
+    - 支持 BF16/FP8 和 MTP。
 
 
 多模态模型
@@ -93,4 +95,4 @@ Reward模型
   * - `internLM-reward <https://huggingface.co/internlm/internlm2-1_8b-reward>`_
     -  :code:`--use_reward_model`
   * - `Qwen2-Reward <https://huggingface.co/Qwen/Qwen2-Reward>`_
-    -  :code:`--use_reward_model`
+    -  :code:`--use_reward_model`

docs/EN/source/models/supported_models.rst

@@ -56,6 +56,8 @@ Large Language Models
     - 
   * - `DeepSeek-V3.2 `_
     - 
+  * - `GLM-5.2 <https://huggingface.co/zai-org/GLM-5.2>`_
+    - Supports BF16/FP8 and MTP.
 
 Multimodal Models
 ^^^^^^^^^^^^^^^^^
@@ -94,4 +96,3 @@ Reward Models
     -  :code:`--use_reward_model`
   * - `Qwen2-Reward <https://huggingface.co/Qwen/Qwen2-Reward>`_
     -  :code:`--use_reward_model`
-

lightllm/common/basemodel/attention/nsa/flashmla_sparse.py

@@ -3,6 +3,7 @@
 
 import dataclasses
 import torch
+import torch.nn.functional as F
 from typing import Tuple, TYPE_CHECKING
 
 from ..base_att import BaseAttBackend, BasePrefillAttState, BaseDecodeAttState, AttControl
@@ -86,14 +87,20 @@ def _nsa_prefill_att(
         if topk_mem_indices.ndim == 2:
             topk_mem_indices = topk_mem_indices.unsqueeze(1)
 
+        real_head_num = q.shape[1]
+        head_block_size = 64
+        pad_head_num = (-real_head_num) % head_block_size
+        if pad_head_num:
+            q = F.pad(q, (0, 0, 0, pad_head_num))
+
         mla_out, _, _ = flash_mla_sparse_fwd(
             q=q,
             kv=kv,
             indices=topk_mem_indices,
             sm_scale=softmax_scale,
             d_v=kv_lora_rank,
         )
-        return mla_out
+        return mla_out[:, :real_head_num, :]
 
 
 @dataclasses.dataclass

lightllm/common/basemodel/attention/nsa/fp8_flashmla_sparse.py

@@ -1,5 +1,6 @@
 import dataclasses
 import torch
+import torch.nn.functional as F
 from typing import TYPE_CHECKING, Tuple
 
 from ..base_att import AttControl, BaseAttBackend, BaseDecodeAttState, BasePrefillAttState
@@ -70,10 +71,9 @@ def _nsa_prefill_att(
         packed_kv: torch.Tensor,
         att_control: AttControl,
     ) -> torch.Tensor:
-        import flash_mla
+        from flash_mla import flash_mla_sparse_fwd
 
         nsa_dict = att_control.nsa_prefill_dict
-        topk_indices = nsa_dict["topk_indices"]
         softmax_scale = nsa_dict["softmax_scale"]
         kv_lora_rank = nsa_dict["kv_lora_rank"]
         topk_mem_indices = nsa_dict["topk_mem_indices"]
@@ -91,18 +91,25 @@ def _nsa_prefill_att(
             )
         else:
             kv = prefill_cache_kv
+            topk_indices = nsa_dict["topk_indices"]
 
         if topk_indices.ndim == 2:
             topk_indices = topk_indices.unsqueeze(1)
 
-        mla_out, _, _ = flash_mla.flash_mla_sparse_fwd(
+        real_head_num = q.shape[1]
+        head_block_size = 64
+        pad_head_num = (-real_head_num) % head_block_size
+        if pad_head_num:
+            q = F.pad(q, (0, 0, 0, pad_head_num))
+
+        mla_out, _, _ = flash_mla_sparse_fwd(
             q=q,
             kv=kv,
             indices=topk_indices,
             sm_scale=softmax_scale,
             d_v=kv_lora_rank,
         )
-        return mla_out
+        return mla_out[:, :real_head_num, :]
 
 
 @dataclasses.dataclass
@@ -141,9 +148,9 @@ def init_state(self):
             ragged_mem_index=self.ragged_mem_index,
             hold_req_idx=self.infer_state.req_manager.HOLD_REQUEST_ID,
         )
-        import flash_mla
+        from flash_mla import get_mla_metadata
 
-        self.flashmla_sched_meta, _ = flash_mla.get_mla_metadata()
+        self.flashmla_sched_meta, _ = get_mla_metadata()
         return
 
     def decode_att(
@@ -164,7 +171,7 @@ def _nsa_decode_att(
         packed_kv: torch.Tensor,
         att_control: AttControl,
     ) -> torch.Tensor:
-        import flash_mla
+        from flash_mla import flash_mla_with_kvcache
 
         nsa_dict = att_control.nsa_decode_dict
         topk_mem_indices = nsa_dict["topk_mem_indices"]
@@ -177,13 +184,26 @@ def _nsa_decode_att(
 
         q_nope, q_rope = q
         q_all = torch.cat([q_nope, q_rope], dim=-1).unsqueeze(1).contiguous()
+
+        real_head_num = q_all.shape[2]
+        if real_head_num <= 64:
+            padded_head_num = 64
+        elif real_head_num <= 128:
+            padded_head_num = 128
+        else:
+            padded_head_num = real_head_num
+        if padded_head_num != real_head_num:
+            q_all = F.pad(q_all, (0, 0, 0, padded_head_num - real_head_num))
+
+        # Token-granular "pages": indices=topk_mem_indices are absolute KV-pool slots,
+        # so the kv cache is viewed as (num_tokens, 1, 1, bytes) and no block_table is needed.
         kv = torch.as_strided(
             packed_kv,
             size=(packed_kv.shape[0], 1, 1, packed_kv.shape[-1]),
             stride=(packed_kv.stride(0), packed_kv.shape[-1], packed_kv.shape[-1], packed_kv.stride(-1)),
         )
 
-        o_tensor, _ = flash_mla.flash_mla_with_kvcache(
+        o_tensor, _ = flash_mla_with_kvcache(
             q=q_all,
             k_cache=kv,
             block_table=None,
@@ -193,6 +213,7 @@ def _nsa_decode_att(
             softmax_scale=softmax_scale,
             causal=False,
             is_fp8_kvcache=True,
-            indices=topk_mem_indices,
+            indices=topk_mem_indices.to(dtype=torch.int32),
         )
+        o_tensor = o_tensor[:, :, :real_head_num, :]
         return o_tensor[:, 0, :, :]  # [b, 1, h, d] -> [b, h, d]

lightllm/common/basemodel/basemodel.py

@@ -1177,12 +1177,7 @@ def _init_padded_req(self):
     def _gen_special_model_input(self, token_num: int):
         special_model_input = {}
 
-        is_mtp_draft_model = (
-            "Deepseek3MTPModel" in str(self.__class__)
-            or "Qwen3MOEMTPModel" in str(self.__class__)
-            or "MistralMTPModel" in str(self.__class__)
-            or "Glm4MoeLiteMTPModel" in str(self.__class__)
-        )
+        is_mtp_draft_model = getattr(self, "is_mtp_draft_model", False)
         if is_mtp_draft_model:
             special_model_input["mtp_draft_input_hiddens"] = torch.randn(
                 token_num, self.config["hidden_size"], dtype=self.data_type, device="cuda"

lightllm/common/basemodel/layer_weights/meta_weights/norm_weight.py

@@ -3,6 +3,7 @@
 from .base_weight import BaseWeightTpl
 from lightllm.utils.dist_utils import get_current_device_id, get_current_rank_in_dp, get_dp_world_size
 from lightllm.common.basemodel.triton_kernel.norm.rmsnorm import rmsnorm_forward
+from lightllm.common.basemodel.triton_kernel.norm.fused_add_rmsnorm import fused_add_rmsnorm_forward
 from lightllm.common.basemodel.triton_kernel.norm.layernorm import layernorm_forward
 from lightllm.common.basemodel.triton_kernel.norm.qk_norm import qk_rmsnorm_fused_forward
 from lightllm.common.basemodel.triton_kernel.norm.gated_rmsnorm import gated_rmsnorm_forward
@@ -71,6 +72,21 @@ def __call__(
     ) -> torch.Tensor:
         return self._forward(input=input, eps=eps, out=out, alloc_func=alloc_func)
 
+    def fused_add_forward(
+        self,
+        residual: torch.Tensor,
+        x: torch.Tensor,
+        eps: float,
+        out: Optional[torch.Tensor] = None,
+        alloc_func=torch.empty,
+    ) -> torch.Tensor:
+        """Fused residual-add + RMSNorm: ``residual <- residual + x`` (in place) and return
+        ``rmsnorm(residual) * weight``. Bit-identical to a plain ``residual.add_(x)`` followed
+        by ``__call__`` but in a single Triton launch. CUDA/MUSA (Triton) only."""
+        if out is None:
+            out = alloc_func(residual.shape, dtype=residual.dtype, device=residual.device)
+        return fused_add_rmsnorm_forward(residual=residual, x=x, weight=self.weight, eps=eps, out=out)
+
 
 class GatedRMSNormWeight(RMSNormWeight):
     def _triton_forward(

lightllm/common/basemodel/triton_kernel/fused_moe/grouped_fused_moe.py

@@ -18,6 +18,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import os
 import torch
 import triton
 import triton.language as tl
@@ -26,6 +27,7 @@
 from lightllm.utils.vllm_utils import vllm_ops
 from lightllm.utils.device_utils import triton_support_tensor_descriptor
 from .moe_silu_and_mul import silu_and_mul_fwd
+from .moe_silu_and_mul_group_quant import silu_and_mul_group_quant_fwd
 from .moe_sum_reduce import moe_sum_reduce
 from lightllm.common.basemodel.triton_kernel.quantization.fp8act_quant_kernel import per_token_group_quant_fp8
 from lightllm.utils.torch_ops_utils import direct_register_custom_op
@@ -35,6 +37,11 @@
 
 logger = init_logger(__name__)
 
+# Fuse the down-projection's per-token-group fp8 quant into the silu_and_mul that produces its
+# input (eliminates a separate per_token_group_quant launch per MoE layer). Numerically matches
+# the unfused path; gate exists for A/B and rollback.
+_ENABLE_FUSED_SILU_QUANT = os.environ.get("LIGHTLLM_FUSED_SILU_QUANT", "1") == "1"
+
 
 @triton.jit
 def moe_align_kernel(
@@ -860,8 +867,12 @@ def grouped_matmul(
         assert BLOCK_SIZE_K == triton.next_power_of_2(BLOCK_SIZE_K)
 
     if use_fp8_w8a8:
+        if token_inputs.dtype == expert_weights.dtype:
+            # token_inputs were already fp8-quantized by a fused producer (e.g. the fused
+            # silu+mul+group-quant on the down projection); token_input_scale is its scale.
+            assert token_input_scale is not None, "pre-quantized fp8 input requires its scale"
         # 当权重使用 block wise 量化时，激活也使用 per token， group size 量化
-        if block_size_k == 0:
+        elif block_size_k == 0:
             # input 使用 per token 量化
             token_inputs, token_input_scale = vllm_ops.scaled_fp8_quant(
                 token_inputs, token_input_scale, use_per_token_if_dynamic=True
@@ -1082,18 +1093,44 @@ def fused_experts_impl(
             bias=w1_bias,
         )
 
-        silu_and_mul_fwd(
-            intermediate_cache1.view(-1, N),
-            intermediate_cache2.view(-1, N // 2),
-            limit=limit,
-            alpha=alpha,
-            layout=layout,
-        )
+        # Fuse the down-projection's per-token-group fp8 quant into silu_and_mul when the down
+        # weight is block-wise fp8 quantized: the silu output is emitted directly as fp8 + scales,
+        # so grouped_matmul below skips its internal per_token_group_quant launch.
+        use_fused_silu_quant = _ENABLE_FUSED_SILU_QUANT and use_fp8_w8a8 and w2_scale is not None and w2_scale.ndim == 3
+        if use_fused_silu_quant:
+            down_token_num = curr_topk_ids.numel()
+            down_k = N // 2
+            down_group_size = down_k // w2_scale.shape[2]
+            down_inputs = alloc_tensor_func(
+                (down_token_num, down_k), device=hidden_states.device, dtype=torch.float8_e4m3fn
+            )
+            down_input_scale = alloc_tensor_func(
+                (down_token_num, down_k // down_group_size), device=hidden_states.device, dtype=torch.float32
+            )
+            silu_and_mul_group_quant_fwd(
+                intermediate_cache1.view(-1, N),
+                down_inputs,
+                down_input_scale,
+                down_group_size,
+                layout=layout,
+                limit=limit,
+                alpha=alpha,
+            )
+        else:
+            silu_and_mul_fwd(
+                intermediate_cache1.view(-1, N),
+                intermediate_cache2.view(-1, N // 2),
+                limit=limit,
+                alpha=alpha,
+                layout=layout,
+            )
+            down_inputs = intermediate_cache2.view(-1, N // 2)
+            down_input_scale = a2_scale
 
         grouped_matmul(
             curr_topk_ids.numel(),
-            intermediate_cache2.view(-1, N // 2),
-            a2_scale,
+            down_inputs,
+            down_input_scale,
             expert_to_token_num,
             expert_to_tokens,
             expert_to_weights=expert_to_weights,