[Feature] Define backends and add Triton backend for Lora (#3161)

Co-authored-by: Ying Sheng <sqy1415@gmail.com>

python/sglang/test/runners.py

@@ -272,6 +272,7 @@ class SRTRunner:
         port: int = DEFAULT_PORT_FOR_SRT_TEST_RUNNER,
         lora_paths: List[str] = None,
         max_loras_per_batch: int = 4,
+        lora_backend: str = "triton",
         disable_cuda_graph: bool = False,
         disable_radix_cache: bool = False,
     ):
@@ -287,6 +288,7 @@ class SRTRunner:
             is_embedding=not self.is_generation,
             lora_paths=lora_paths,
             max_loras_per_batch=max_loras_per_batch,
+            lora_backend=lora_backend,
             disable_cuda_graph=disable_cuda_graph,
             disable_radix_cache=disable_radix_cache,
         )