[Doc] ruff format remaining Python examples (#26795)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

docs/serving/integrations/langchain.md

@@ -15,13 +15,15 @@ To run inference on a single or multiple GPUs, use `VLLM` class from `langchain`
     ```python
     from langchain_community.llms import VLLM
 
-    llm = VLLM(model="mosaicml/mpt-7b",
-            trust_remote_code=True,  # mandatory for hf models
-            max_new_tokens=128,
-            top_k=10,
-            top_p=0.95,
-            temperature=0.8,
-            # tensor_parallel_size=... # for distributed inference
+    llm = VLLM(
+        model="mosaicml/mpt-7b",
+        trust_remote_code=True,  # mandatory for hf models
+        max_new_tokens=128,
+        top_k=10,
+        top_p=0.95,
+        temperature=0.8,
+        # for distributed inference
+        # tensor_parallel_size=...,
     )
 
     print(llm("What is the capital of France ?"))