patch: remove usage of llm, chat model __call__ (#20788)

- `llm(prompt)` -> `llm.invoke(prompt)`
- `llm(prompt=prompt` -> `llm.invoke(prompt)` (same with `messages=`)
- `llm(prompt, callbacks=callbacks)` -> `llm.invoke(prompt,
config={"callbacks": callbacks})`
- `llm(prompt, **kwargs)` -> `llm.invoke(prompt, **kwargs)`

libs/community/tests/integration_tests/llms/test_ai21.py

@@ -9,14 +9,14 @@ from langchain_community.llms.loading import load_llm
 def test_ai21_call() -> None:
     """Test valid call to ai21."""
     llm = AI21(maxTokens=10)
-    output = llm("Say foo:")
+    output = llm.invoke("Say foo:")
     assert isinstance(output, str)
 
 
 def test_ai21_call_experimental() -> None:
     """Test valid call to ai21 with an experimental model."""
     llm = AI21(maxTokens=10, model="j1-grande-instruct")
-    output = llm("Say foo:")
+    output = llm.invoke("Say foo:")
     assert isinstance(output, str)