Refactor vllm_hpu -> vllm_gaudi

docs/.nav.yml

@@ -33,5 +33,5 @@ nav:
   - API Reference:
     - Summary: api/README.md
     - Contents:
-      - glob: api/vllm_hpu/*
+      - glob: api/vllm_gaudi/*
         preserve_directory_names: true

mkdocs.yaml

@@ -55,7 +55,7 @@ plugins:
   - awesome-nav
   # For API reference generation
   - api-autonav:
-      modules: ["vllm_hpu"]
+      modules: ["vllm_gaudi"]
       api_root_uri: "api"
   - mkdocstrings:
       handlers:

vllm_hpu/__init__.py → vllm_gaudi/__init__.py

@@ -1,12 +1,12 @@
-from vllm_hpu.platform import HpuPlatform
+from vllm_gaudi.platform import HpuPlatform
 
 
 def register():
     """Register the HPU platform."""
     HpuPlatform.set_torch_compile()
-    return "vllm_hpu.platform.HpuPlatform"
+    return "vllm_gaudi.platform.HpuPlatform"
 
 
 def register_ops():
     """Register custom ops for the HPU platform."""
-    import vllm_hpu.ops  # noqa: F401
+    import vllm_gaudi.ops  # noqa: F401

vllm_hpu/extension/test_bucketing.py → vllm_gaudi/extension/test_bucketing.py

@@ -8,8 +8,8 @@
 
 import pytest
 
-import vllm_hpu.extension.bucketing.linear as linear
-from vllm_hpu.extension.bucketing.exponential import HPUExponentialBucketingContext
+import vllm_gaudi.extension.bucketing.linear as linear
+from vllm_gaudi.extension.bucketing.exponential import HPUExponentialBucketingContext
 
 
 @pytest.fixture

vllm_hpu/platform.py → vllm_gaudi/platform.py

@@ -64,7 +64,7 @@ def check_and_update_config(cls, vllm_config: VllmConfig) -> None:
         if parallel_config.worker_cls == "auto":
             if envs.VLLM_USE_V1:
                 parallel_config.worker_cls = \
-                    "vllm_hpu.v1.worker.hpu_worker.HPUWorker"
+                    "vllm_gaudi.v1.worker.hpu_worker.HPUWorker"
             else:
                 parallel_config.worker_cls = \
                     "vllm.worker.hpu_worker.HPUWorker"
@@ -117,11 +117,11 @@ def is_pin_memory_available(cls):
 
     @classmethod
     def get_punica_wrapper(cls) -> str:
-        return "vllm_hpu.lora.punica_wrapper.punica_hpu.PunicaWrapperHPU"
+        return "vllm_gaudi.lora.punica_wrapper.punica_hpu.PunicaWrapperHPU"
 
     @classmethod
     def get_device_communicator_cls(cls) -> str:
-        return "vllm_hpu.distributed.device_communicators.hpu_communicator.HpuCommunicator"  # noqa
+        return "vllm_gaudi.distributed.device_communicators.hpu_communicator.HpuCommunicator"  # noqa
 
     @classmethod
     def supports_structured_output(cls) -> bool: