{"record":{"id":"83e58f49d8b8dc34","repo":"vllm-project/vllm","slug":"eplb-communicator-pynccl-requested-but-expert-we","errorCode":null,"errorMessage":"EPLB communicator 'pynccl' requested but expert weights contain unsupported dtypes: ({', '.join(str(dtype) for dtype in unsupported_dtypes)}).","messagePattern":"EPLB communicator 'pynccl' requested but expert weights contain unsupported dtypes: \\((.+?)\\)\\.","errorType":"exception","errorClass":"RuntimeError","httpStatus":null,"severity":"error","filePath":"vllm/distributed/eplb/eplb_communicator.py","lineNumber":708,"sourceCode":"        else group_coordinator.device_group\n    )\n\n    def _create_pynccl() -> EplbCommunicator:\n        if tensor_device_type == \"cpu\":\n            raise RuntimeError(\n                \"EPLB communicator 'pynccl' supports only cuda-like devices \"\n                f\"(got {tensor_device_type}).\"\n            )\n        unsupported_dtypes = sorted(\n            {\n                tensor.dtype\n                for tensor in first_layer\n                if not ncclDataTypeEnum.supports_torch_dtype(tensor.dtype)\n            },\n            key=str,\n        )\n        if unsupported_dtypes:\n            raise RuntimeError(\n                \"EPLB communicator 'pynccl' requested but expert weights contain \"\n                \"unsupported dtypes: \"\n                f\"({', '.join(str(dtype) for dtype in unsupported_dtypes)}).\"\n            )\n\n        device_comm = group_coordinator.device_communicator\n        pynccl_comm = (\n            getattr(device_comm, \"pynccl_comm\", None)\n            if device_comm is not None\n            else None\n        )\n        if pynccl_comm is None or pynccl_comm.disabled or not pynccl_comm.available:\n            raise RuntimeError(\"EPLB communicator 'pynccl' requested but unavailable.\")\n        try:\n            return PyNcclEplbCommunicator(pynccl_comm=pynccl_comm)\n        except Exception as exc:\n            raise RuntimeError(\n                f\"Failed to initialize PyNcclEplbCommunicator ({exc}).\"","sourceCodeStart":690,"sourceCodeEnd":726,"githubUrl":"https://github.com/vllm-project/vllm/blob/c794754062d49a8fdb63ab3c5215b488b865030c/vllm/distributed/eplb/eplb_communicator.py#L690-L726","documentation":"Error \"EPLB communicator 'pynccl' requested but expert weights contain unsupported dtypes: ({', '.join(str(dtype) for dtype in unsupported_dtypes)}).\" thrown in vllm-project/vllm.","triggerScenarios":"Raised at vllm/distributed/eplb/eplb_communicator.py:708 when validation fails: EPLB communicator 'pynccl' requested but expert weights contain unsupported dtypes. Typically triggered by an incompatible or incomplete vLLM configuration, an unsupported platform/backend combination, or a runtime resource/dependency that is missing.","commonSituations":"Commonly encountered at vllm/distributed/eplb/eplb_communicator.py:708 during vLLM startup/config validation or runtime setup when: (1) conflicting CLI flags or config fields are combined, (2) the current platform (CUDA/ROCm/CPU/XPU) or installed optional packages do not support the requested feature, or (3) a required value is absent or out of range. Resolve by correcting the configuration as described in the message, or by selecting a supported alternative.","solutions":["Convert expert weights to dtypes supported by pynccl, or choose a different EPLB communicator backend."],"exampleFix":null,"handlingStrategy":null,"validationCode":null,"typeGuard":null,"tryCatchPattern":null,"preventionTips":[],"tags":[],"backgroundTag":null,"analyzedSha":"c794754062d49a8fdb63ab3c5215b488b865030c","analyzedAt":"2026-08-14T21:17:39.825Z","schemaVersion":2},"datasetVersion":"2026-08-15T22:17:37.221Z"}