fix vllmsampler client (#122)

Author: tastelikefeet

client_tools/client_generator.py

@@ -805,6 +805,14 @@ def set_template(self, template_cls: str, adapter_name: str = '', **kwargs) -> S
         )
         response.raise_for_status()
         return SetTemplateResponse(**response.json())
+    
+    def apply_patch(self, patch_cls: str, **kwargs) -> None:
+        """Apply a patch to the model."""
+        response = http_post(
+            url=f'{self.server_url}/apply_patch',
+            json_data={'patch_cls': patch_cls, 'adapter_name': self.adapter_name, **kwargs}
+        )
+        response.raise_for_status()
 '''
 
     # Write the sampler client file

src/twinkle/server/sampler/twinkle_handlers.py

@@ -8,7 +8,9 @@
 
 import traceback
 from fastapi import Depends, FastAPI, HTTPException, Request
-from typing import TYPE_CHECKING, Callable, List, Optional
+from typing import TYPE_CHECKING, Callable
+
+from twinkle.server.common.serialize import deserialize_object
 
 if TYPE_CHECKING:
     from .app import SamplerManagement
@@ -162,3 +164,13 @@ def add_adapter_to_sampler(
         self.sampler.add_adapter_to_sampler(full_adapter_name, config)
 
         return types.AddAdapterResponse(adapter_name=full_adapter_name)
+
+    @app.post('/twinkle/apply_patch')
+    async def apply_patch(
+            request: Request,
+            body: types.ApplyPatchRequest,
+            self: SamplerManagement = Depends(self_fn),
+    ) -> None:
+        extra_kwargs = body.model_extra or {}
+        patch_cls = deserialize_object(body.patch_cls)
+        self.sampler.apply_patch(patch_cls, **extra_kwargs)

src/twinkle_client/sampler/vllm_sampler.py

@@ -94,3 +94,11 @@ def set_template(self, template_cls: str, adapter_name: str = '', **kwargs) -> S
         )
         response.raise_for_status()
         return SetTemplateResponse(**response.json())
+
+    def apply_patch(self, patch_cls: str, **kwargs) -> None:
+        """Apply a patch to the model."""
+        response = http_post(
+            url=f'{self.server_url}/apply_patch',
+            json_data={'patch_cls': patch_cls, 'adapter_name': self.adapter_name, **kwargs}
+        )
+        response.raise_for_status()