[ROCm][Bugfix] Explicitly reject FSE=1 with DPA+ETP deployment for DeepSeek-V4

tests/models/test_deepseek_v4_vl_rocm.py

@@ -7,6 +7,7 @@
 import torch
 from torch import nn
 
+from vllm.config import ParallelConfig
 from vllm.model_executor.models.utils import WeightsMapper
 from vllm.platforms import current_platform
 
@@ -100,7 +101,9 @@ def fake_factory(**kwargs):
         n_shared_experts=None,
     )
     vllm_config = SimpleNamespace(
-        model_config=SimpleNamespace(hf_config=config), quant_config=None
+        model_config=SimpleNamespace(hf_config=config),
+        quant_config=None,
+        parallel_config=ParallelConfig(),
     )
 
     hash_moe = rocm_model.DeepseekV4MoE(vllm_config, prefix="model.layers.0.ffn")
@@ -115,6 +118,24 @@ def fake_factory(**kwargs):
         assert factory_kwargs["image_sentinel_lo"] == IMAGE_SENTINEL_BASE_ID
 
 
+def test_rocm_fse_rejects_data_parallel_deployment(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    from vllm.models.deepseek_v4.amd.model import _fuse_shared_experts_enabled
+
+    config = SimpleNamespace(n_shared_experts=1)
+
+    monkeypatch.setenv("VLLM_ROCM_USE_AITER_FUSION_SHARED_EXPERTS", "1")
+    with pytest.raises(ValueError, match="data parallelism"):
+        _fuse_shared_experts_enabled(config, ParallelConfig(data_parallel_size=2))
+
+    # Expert parallelism already disables FSE: no raise, FSE stays off.
+    assert not _fuse_shared_experts_enabled(
+        config, ParallelConfig(data_parallel_size=2, enable_expert_parallel=True)
+    )
+    assert _fuse_shared_experts_enabled(config, ParallelConfig())
+
+
 def test_rocm_mtp_forwards_input_ids_for_vision_routing(
     monkeypatch: pytest.MonkeyPatch,
 ) -> None:

vllm/models/deepseek_v4/amd/model.py

@@ -12,7 +12,7 @@
 
 import vllm.envs as envs
 from vllm._aiter_ops import rocm_aiter_ops
-from vllm.config import VllmConfig, get_current_vllm_config
+from vllm.config import ParallelConfig, VllmConfig
 from vllm.distributed import (
     get_pp_group,
     get_tensor_model_parallel_rank,
@@ -504,11 +504,28 @@ def forward_modular(
         return routed + shared_out
 
 
-def _fuse_shared_experts_enabled(config) -> bool:
+def _fuse_shared_experts_enabled(config, parallel_config: ParallelConfig) -> bool:
+    if (
+        getattr(config, "n_shared_experts", None)
+        and envs.VLLM_ROCM_USE_AITER_FUSION_SHARED_EXPERTS
+        and not parallel_config.enable_expert_parallel
+        and parallel_config.data_parallel_size > 1
+    ):
+        # Fused shared experts are not supported under data parallelism:
+        # the fused path cannot load the FP8 shared expert into the MXFP4
+        # routed slot, which fails obscurely during weight loading. Fail
+        # fast with an actionable message instead.
+        raise ValueError(
+            "DeepSeek-V4 fused shared experts "
+            "(VLLM_ROCM_USE_AITER_FUSION_SHARED_EXPERTS=1) are not supported "
+            "with data parallelism (data_parallel_size > 1). Set "
+            "VLLM_ROCM_USE_AITER_FUSION_SHARED_EXPERTS=0 to run the shared "
+            "expert as an unfused FP8 MLP."
+        )
     return bool(
         getattr(config, "n_shared_experts", None)
         and envs.VLLM_ROCM_USE_AITER_FUSION_SHARED_EXPERTS
-        and not get_current_vllm_config().parallel_config.enable_expert_parallel
+        and not parallel_config.enable_expert_parallel
     )
 
 
@@ -586,7 +603,7 @@ def __init__(
         # This should be cleaned up and use `resolve_layer_fused_shared_expert`.
         self.fuse_heterogeneous_shared_expert = fuse_heterogeneous_shared_expert
         fse_requested = (
-            _fuse_shared_experts_enabled(config)
+            _fuse_shared_experts_enabled(config, vllm_config.parallel_config)
             and not self.fuse_heterogeneous_shared_expert
         )
         fse_compatible = False