Merge tag 'v0.6.2' into v0.6.2-dev

vllm/worker/utils.py

@@ -39,18 +39,10 @@ def assert_enc_dec_mr_supported_scenario(
         raise NotImplementedError(
             STR_NOT_IMPL_ENC_DEC_ERR_STRS['STR_NOT_IMPL_ENC_DEC_PP'])
 
-    if enc_dec_mr.model_config.is_multimodal_model:
-        raise NotImplementedError(
-            STR_NOT_IMPL_ENC_DEC_ERR_STRS['STR_NOT_IMPL_ENC_DEC_MM'])
-
     if enc_dec_mr.scheduler_config.num_lookahead_slots > 0:
         raise NotImplementedError(
             STR_NOT_IMPL_ENC_DEC_ERR_STRS['STR_NOT_IMPL_ENC_DEC_SPEC_DEC'])
 
-    if not enc_dec_mr.model_config.enforce_eager:
-        raise NotImplementedError(
-            STR_NOT_IMPL_ENC_DEC_ERR_STRS['STR_NOT_IMPL_ENC_DEC_CUDA_GRAPH'])
-
     if enc_dec_mr.prompt_adapter_config is not None:
         raise NotImplementedError(STR_NOT_IMPL_ENC_DEC_ERR_STRS[
             'STR_NOT_IMPL_ENC_DEC_PROMPT_ADAPTER'])

vllm/worker/worker.py

@@ -454,14 +454,20 @@ def init_worker_distributed_environment(
 
 def _check_if_gpu_supports_dtype(torch_dtype: torch.dtype):
     # Check if the GPU supports the dtype.
-    if torch_dtype == torch.bfloat16:
-        compute_capability = current_platform.get_device_capability()
-        if compute_capability[0] < 8:
+    if torch_dtype == torch.bfloat16:  # noqa: SIM102
+        if not current_platform.has_device_capability(80):
+            capability = current_platform.get_device_capability()
             gpu_name = current_platform.get_device_name()
+
+            if capability is None:
+                compute_str = "does not have a compute capability"
+            else:
+                version_str = capability.as_version_str()
+                compute_str = f"has compute capability {version_str}"
+
             raise ValueError(
                 "Bfloat16 is only supported on GPUs with compute capability "
-                f"of at least 8.0. Your {gpu_name} GPU has compute capability "
-                f"{compute_capability[0]}.{compute_capability[1]}. "
+                f"of at least 8.0. Your {gpu_name} GPU {compute_str}. "
                 "You can use float16 instead by explicitly setting the"
                 "`dtype` flag in CLI, for example: --dtype=half.")
 

vllm/worker/xpu_model_runner.py

@@ -21,7 +21,7 @@ from vllm.multimodal import (MULTIMODAL_REGISTRY, BatchedTensorInputs,
                              MultiModalInputs, MultiModalRegistry)
 from vllm.sampling_params import SamplingParams
 from vllm.sequence import IntermediateTensors, SequenceGroupMetadata
-from vllm.utils import CudaMemoryProfiler, make_tensor_with_pad
+from vllm.utils import DeviceMemoryProfiler, make_tensor_with_pad
 from vllm.worker.model_runner import AttentionMetadata, SamplingMetadata
 from vllm.worker.model_runner_base import (
     ModelRunnerBase, ModelRunnerInputBase, ModelRunnerInputBuilderBase,
@@ -391,7 +391,7 @@ class XPUModelRunner(ModelRunnerBase[ModelInputForXPUWithSamplingMetadata]):
         self.model: nn.Module  # Set after init_Model
 
     def load_model(self) -> None:
-        with CudaMemoryProfiler() as m:
+        with DeviceMemoryProfiler() as m:
             self.model = get_model(
                 model_config=self.model_config,
                 device_config=self.device_config,