feat: add lazy LLM initialization in RankGPTRerank (#10648)

llama-index-integrations/postprocessor/llama-index-postprocessor-rankgpt-rerank/llama_index/postprocessor/rankgpt_rerank/base.py

@@ -9,7 +9,6 @@ from llama_index.core.prompts.default_prompts import RANKGPT_RERANK_PROMPT
 from llama_index.core.prompts.mixin import PromptDictType
 from llama_index.core.schema import NodeWithScore, QueryBundle
 from llama_index.core.utils import print_text
-from llama_index.llms.openai import OpenAI
 
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.WARNING)
@@ -19,10 +18,7 @@ class RankGPTRerank(BaseNodePostprocessor):
     """RankGPT-based reranker."""
 
     top_n: int = Field(default=5, description="Top N nodes to return from reranking.")
-    llm: LLM = Field(
-        default_factory=lambda: OpenAI(model="gpt-3.5-turbo-16k"),
-        description="LLM to use for rankGPT",
-    )
+    llm: Optional[LLM] = None
     verbose: bool = Field(
         default=False, description="Whether to print intermediate steps."
     )
@@ -49,6 +45,18 @@ class RankGPTRerank(BaseNodePostprocessor):
     def class_name(cls) -> str:
         return "RankGPTRerank"
 
+    def _ensure_llm(self) -> None:
+        if not self.llm:
+            try:
+                from llama_index.llms.openai import OpenAI
+
+                self.llm = OpenAI(model="gpt-3.5-turbo-16k")
+            except ImportError:
+                raise RuntimeError(
+                    "OpenAI LLM is not available. Please install `llama-index-llms-openai` "
+                    "or provide an alternative LLM instance."
+                )
+
     def _postprocess_nodes(
         self,
         nodes: List[NodeWithScore],
@@ -129,6 +137,7 @@ class RankGPTRerank(BaseNodePostprocessor):
         return messages
 
     def run_llm(self, messages: Sequence[ChatMessage]) -> ChatResponse:
+        self._ensure_llm()
         return self.llm.chat(messages)
 
     def _clean_response(self, response: str) -> str: