community[patch]: update use of deprecated llm methods (#20393)

.predict and .predict_messages for BaseLanguageModel and BaseChatModel

libs/community/langchain_community/callbacks/labelstudio_callback.py

@@ -98,7 +98,7 @@ class LabelStudioCallbackHandler(BaseCallbackHandler):
         ...             mode='prompt'
         ... )
         >>> llm = OpenAI(callbacks=[handler])
-        >>> llm.predict('Tell me a story about a dog.')
+        >>> llm.invoke('Tell me a story about a dog.')
     """
 
     DEFAULT_PROJECT_NAME: str = "LangChain-%Y-%m-%d"

libs/community/langchain_community/callbacks/llmonitor_callback.py

@@ -204,7 +204,7 @@ class LLMonitorCallbackHandler(BaseCallbackHandler):
     llmonitor_callback = LLMonitorCallbackHandler()
     llm = OpenAI(callbacks=[llmonitor_callback],
                  metadata={"userId": "user-123"})
-    llm.predict("Hello, how are you?")
+    llm.invoke("Hello, how are you?")
     ```
     """
 

libs/community/langchain_community/llms/opaqueprompts.py

@@ -4,6 +4,7 @@ from typing import Any, Dict, List, Optional
 from langchain_core.callbacks import CallbackManagerForLLMRun
 from langchain_core.language_models import BaseLanguageModel
 from langchain_core.language_models.llms import LLM
+from langchain_core.messages import AIMessage
 from langchain_core.pydantic_v1 import Extra, root_validator
 from langchain_core.utils import get_from_dict_or_env
 
@@ -95,10 +96,11 @@ class OpaquePrompts(LLM):
 
         # TODO: Add in callbacks once child runs for LLMs are supported by LangSmith.
         # call the LLM with the sanitized prompt and get the response
-        llm_response = self.base_llm.predict(
+        llm_response = self.base_llm.bind(stop=stop).invoke(
             sanitized_prompt_value_str,
-            stop=stop,
         )
+        if isinstance(llm_response, AIMessage):
+            llm_response = llm_response.content
 
         # desanitize the response by restoring the original sensitive information
         desanitize_response: op.DesanitizeResponse = op.desanitize(

libs/community/tests/unit_tests/chat_models/test_openai.py

@@ -96,8 +96,8 @@ def test_openai_predict(mock_completion: dict) -> None:
         "client",
         mock_client,
     ):
-        res = llm.predict("bar")
+        res = llm.invoke("bar")
-        assert res == "Bar Baz"
+        assert res.content == "Bar Baz"
     assert completed
 
 

libs/core/tests/unit_tests/language_models/llms/test_base.py

@@ -60,7 +60,7 @@ def test_batch_size() -> None:
 
     llm = FakeListLLM(responses=["foo"] * 1)
     with collect_runs() as cb:
-        llm.predict("foo")
+        llm.invoke("foo")
         assert len(cb.traced_runs) == 1
         assert (cb.traced_runs[0].extra or {}).get("batch_size") == 1
 

libs/core/tests/unit_tests/tracers/test_run_collector.py

@@ -9,7 +9,7 @@ from langchain_core.tracers.context import collect_runs
 def test_collect_runs() -> None:
     llm = FakeListLLM(responses=["hello"])
     with collect_runs() as cb:
-        llm.predict("hi")
+        llm.invoke("hi")
         assert cb.traced_runs
         assert len(cb.traced_runs) == 1
         assert isinstance(cb.traced_runs[0].id, uuid.UUID)

libs/experimental/langchain_experimental/llms/anthropic_functions.py

@@ -183,7 +183,7 @@ class AnthropicFunctions(BaseChatModel):
                 raise ValueError(
                     "if `function_call` provided, `functions` must also be"
                 )
-        response = self.model.predict_messages(
+        response = self.model.invoke(
             messages, stop=stop, callbacks=run_manager, **kwargs
         )
         completion = cast(str, response.content)

libs/experimental/langchain_experimental/llms/ollama_functions.py

@@ -89,7 +89,7 @@ function in "functions".'
         )
         if "functions" in kwargs:
             del kwargs["functions"]
-        response_message = self.llm.predict_messages(
+        response_message = self.llm.invoke(
             [system_message] + messages, stop=stop, callbacks=run_manager, **kwargs
         )
         chat_generation_content = response_message.content

libs/experimental/tests/unit_tests/chat_models/test_llm_wrapper_llama2chat.py

@@ -49,7 +49,7 @@ def model_cfg_sys_msg() -> Llama2Chat:
 def test_default_system_message(model: Llama2Chat) -> None:
     messages = [HumanMessage(content="usr-msg-1")]
 
-    actual = model.predict_messages(messages).content  # type: ignore
+    actual = model.invoke(messages).content  # type: ignore
     expected = (
         f"<s>[INST] <<SYS>>\n{DEFAULT_SYSTEM_PROMPT}\n<</SYS>>\n\nusr-msg-1 [/INST]"
     )
@@ -62,7 +62,7 @@ def test_configured_system_message(
 ) -> None:
     messages = [HumanMessage(content="usr-msg-1")]
 
-    actual = model_cfg_sys_msg.predict_messages(messages).content  # type: ignore
+    actual = model_cfg_sys_msg.invoke(messages).content  # type: ignore
     expected = "<s>[INST] <<SYS>>\nsys-msg\n<</SYS>>\n\nusr-msg-1 [/INST]"
 
     assert actual == expected
@@ -73,7 +73,7 @@ async def test_configured_system_message_async(
 ) -> None:
     messages = [HumanMessage(content="usr-msg-1")]
 
-    actual = await model_cfg_sys_msg.apredict_messages(messages)  # type: ignore
+    actual = await model_cfg_sys_msg.ainvoke(messages)  # type: ignore
     expected = "<s>[INST] <<SYS>>\nsys-msg\n<</SYS>>\n\nusr-msg-1 [/INST]"
 
     assert actual.content == expected
@@ -87,7 +87,7 @@ def test_provided_system_message(
         HumanMessage(content="usr-msg-1"),
     ]
 
-    actual = model_cfg_sys_msg.predict_messages(messages).content
+    actual = model_cfg_sys_msg.invoke(messages).content
     expected = "<s>[INST] <<SYS>>\ncustom-sys-msg\n<</SYS>>\n\nusr-msg-1 [/INST]"
 
     assert actual == expected
@@ -102,7 +102,7 @@ def test_human_ai_dialogue(model_cfg_sys_msg: Llama2Chat) -> None:
         HumanMessage(content="usr-msg-3"),
     ]
 
-    actual = model_cfg_sys_msg.predict_messages(messages).content
+    actual = model_cfg_sys_msg.invoke(messages).content
     expected = (
         "<s>[INST] <<SYS>>\nsys-msg\n<</SYS>>\n\nusr-msg-1 [/INST] ai-msg-1 </s>"
         "<s>[INST] usr-msg-2 [/INST] ai-msg-2 </s><s>[INST] usr-msg-3 [/INST]"
@@ -113,14 +113,14 @@ def test_human_ai_dialogue(model_cfg_sys_msg: Llama2Chat) -> None:
 
 def test_no_message(model: Llama2Chat) -> None:
     with pytest.raises(ValueError) as info:
-        model.predict_messages([])
+        model.invoke([])
 
     assert info.value.args[0] == "at least one HumanMessage must be provided"
 
 
 def test_ai_message_first(model: Llama2Chat) -> None:
     with pytest.raises(ValueError) as info:
-        model.predict_messages([AIMessage(content="ai-msg-1")])
+        model.invoke([AIMessage(content="ai-msg-1")])
 
     assert (
         info.value.args[0]
@@ -136,7 +136,7 @@ def test_human_ai_messages_not_alternating(model: Llama2Chat) -> None:
     ]
 
     with pytest.raises(ValueError) as info:
-        model.predict_messages(messages)  # type: ignore
+        model.invoke(messages)  # type: ignore
 
     assert info.value.args[0] == (
         "messages must be alternating human- and ai-messages, "
@@ -151,6 +151,6 @@ def test_last_message_not_human_message(model: Llama2Chat) -> None:
     ]
 
     with pytest.raises(ValueError) as info:
-        model.predict_messages(messages)
+        model.invoke(messages)
 
     assert info.value.args[0] == "last message must be a HumanMessage"

libs/experimental/tests/unit_tests/chat_models/test_llm_wrapper_mixtral.py

@@ -23,7 +23,7 @@ def test_prompt(model: Mixtral) -> None:
         HumanMessage(content="usr-msg-2"),
     ]
 
-    actual = model.predict_messages(messages).content  # type: ignore
+    actual = model.invoke(messages).content  # type: ignore
     expected = (
         "<s>[INST] sys-msg\nusr-msg-1 [/INST] ai-msg-1 </s> [INST] usr-msg-2 [/INST]"  # noqa: E501
     )

libs/experimental/tests/unit_tests/chat_models/test_llm_wrapper_orca.py

@@ -23,7 +23,7 @@ def test_prompt(model: Orca) -> None:
         HumanMessage(content="usr-msg-2"),
     ]
 
-    actual = model.predict_messages(messages).content  # type: ignore
+    actual = model.invoke(messages).content  # type: ignore
     expected = "### System:\nsys-msg\n\n### User:\nusr-msg-1\n\n### Assistant:\nai-msg-1\n\n### User:\nusr-msg-2\n\n"  # noqa: E501
 
     assert actual == expected

libs/experimental/tests/unit_tests/chat_models/test_llm_wrapper_vicuna.py

@@ -23,7 +23,7 @@ def test_prompt(model: Vicuna) -> None:
         HumanMessage(content="usr-msg-2"),
     ]
 
-    actual = model.predict_messages(messages).content  # type: ignore
+    actual = model.invoke(messages).content  # type: ignore
     expected = "sys-msg USER: usr-msg-1 ASSISTANT: ai-msg-1 </s>USER: usr-msg-2 "
 
     assert actual == expected

libs/langchain/tests/integration_tests/cache/test_upstash_redis_cache.py

@@ -86,6 +86,5 @@ def test_redis_cache_chat() -> None:
     llm = FakeChatModel()
     params = llm.dict()
     params["stop"] = None
-    with pytest.warns():
+    llm.invoke("foo")
-        llm.predict("foo")
     langchain.llm_cache.redis.flushall()