Support radix cache for Lora feature (#7216)

2025-08-11 10:14:11 -07:00
parent 6f81a710f7
commit 75e6a7cde1
12 changed files with 546 additions and 27 deletions
--- a/test/srt/lora/test_lora_update.py
+++ b/test/srt/lora/test_lora_update.py
@@ -787,7 +787,6 @@ class LoRAUpdateEngineTestSession(LoRAUpdateTestSessionBase):
            max_loaded_loras=self.max_loaded_loras,
            disable_cuda_graph=self.disable_cuda_graph,
            cuda_graph_max_bs=self.cuda_graph_max_bs,
-            disable_radix_cache=True,
            enable_lora=self.enable_lora,
        )
        self.handle.__enter__()
@@ -917,7 +916,6 @@ class LoRAUpdateServerTestSession(LoRAUpdateTestSessionBase):
            str(self.max_loras_per_batch),
            "--lora-backend",
            self.lora_backend,
-            "--disable-radix-cache",
            "--random-seed",
            "42",
            "--max-running-request",