[LoRA] Add support for pinning lora adapters in the LRU cache (#5603)

2024-06-21 15:42:46 -07:00
parent 7187507301
commit f5dda63eb5
13 changed files with 171 additions and 5 deletions
--- a/vllm/executor/cpu_executor.py
+++ b/vllm/executor/cpu_executor.py
@@ -84,6 +84,9 @@ class CPUExecutor(ExecutorBase):
    def remove_lora(self, lora_id: int) -> bool:
        return self.driver_worker.remove_lora(lora_id)

+    def pin_lora(self, lora_id: int) -> bool:
+        return self.driver_worker.pin_lora(lora_id)
+
    def list_loras(self) -> Set[int]:
        return self.driver_worker.list_loras()