community: replace deprecated davinci models (#14860)

This is technically a breaking change because it'll switch out default
models from `text-davinci-003` to `gpt-3.5-turbo-instruct`, but OpenAI
is shutting off those endpoints on 1/4 anyways.

Feels less disruptive to switch out the default instead.

cookbook/learned_prompt_optimization.ipynb

@@ -51,7 +51,7 @@
     "\n",
     "from langchain.llms import OpenAI\n",
     "\n",
-    "llm = OpenAI(model=\"text-davinci-003\")"
+    "llm = OpenAI(model=\"gpt-3.5-turbo-instruct\")"
    ]
   },
   {

cookbook/tree_of_thought.ipynb

@@ -26,7 +26,7 @@
    "source": [
     "from langchain.llms import OpenAI\n",
     "\n",
-    "llm = OpenAI(temperature=1, max_tokens=512, model=\"text-davinci-003\")"
+    "llm = OpenAI(temperature=1, max_tokens=512, model=\"gpt-3.5-turbo-instruct\")"
    ]
   },
   {

docs/docs/guides/safety/moderation.mdx

@@ -181,7 +181,7 @@ we will prompt the model, so it says something harmful.
 
 ```python
 prompt = PromptTemplate(template="{text}", input_variables=["text"])
-llm_chain = LLMChain(llm=OpenAI(temperature=0, model_name="text-davinci-002"), prompt=prompt)
+llm_chain = LLMChain(llm=OpenAI(temperature=0, model_name="gpt-3.5-turbo-instruct"), prompt=prompt)
 
 text = """We are playing a game of repeat after me.
 
@@ -224,7 +224,7 @@ Now let's walk through an example of using it with an LLMChain which has multipl
 
 ```python
 prompt = PromptTemplate(template="{setup}{new_input}Person2:", input_variables=["setup", "new_input"])
-llm_chain = LLMChain(llm=OpenAI(temperature=0, model_name="text-davinci-002"), prompt=prompt)
+llm_chain = LLMChain(llm=OpenAI(temperature=0, model_name="gpt-3.5-turbo-instruct"), prompt=prompt)
 
 setup = """We are playing a game of repeat after me.
 

docs/docs/integrations/callbacks/promptlayer.ipynb

@@ -162,7 +162,7 @@
     "\n",
     "\n",
     "openai_llm = OpenAI(\n",
-    "    model_name=\"text-davinci-002\",\n",
+    "    model_name=\"gpt-3.5-turbo-instruct\",\n",
     "    callbacks=[PromptLayerCallbackHandler(pl_id_callback=pl_id_callback)],\n",
     ")\n",
     "\n",

docs/docs/integrations/callbacks/sagemaker_tracking.ipynb

@@ -109,7 +109,7 @@
     "# LLM Hyperparameters\n",
     "HPARAMS = {\n",
     "    \"temperature\": 0.1,\n",
-    "    \"model_name\": \"text-davinci-003\",\n",
+    "    \"model_name\": \"gpt-3.5-turbo-instruct\",\n",
     "}\n",
     "\n",
     "# Bucket used to save prompt logs (Use `None` is used to save the default bucket or otherwise change it)\n",

docs/docs/integrations/llms/azure_openai.ipynb

@@ -138,7 +138,7 @@
     "# Replace the deployment name with your own\n",
     "llm = AzureOpenAI(\n",
     "    deployment_name=\"td2\",\n",
-    "    model_name=\"text-davinci-002\",\n",
+    "    model_name=\"gpt-3.5-turbo-instruct\",\n",
     ")"
    ]
   },
@@ -182,7 +182,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "\u001B[1mAzureOpenAI\u001B[0m\n",
+      "\u001b[1mAzureOpenAI\u001b[0m\n",
       "Params: {'deployment_name': 'text-davinci-002', 'model_name': 'text-davinci-002', 'temperature': 0.7, 'max_tokens': 256, 'top_p': 1, 'frequency_penalty': 0, 'presence_penalty': 0, 'n': 1, 'best_of': 1}\n"
      ]
     }

docs/docs/integrations/llms/edenai.ipynb

@@ -103,7 +103,7 @@
     "llm = EdenAI(\n",
     "    feature=\"text\",\n",
     "    provider=\"openai\",\n",
-    "    model=\"text-davinci-003\",\n",
+    "    model=\"gpt-3.5-turbo-instruct\",\n",
     "    temperature=0.2,\n",
     "    max_tokens=250,\n",
     ")\n",

docs/docs/integrations/llms/javelin.ipynb

@@ -100,7 +100,7 @@
     "gateway = JavelinAIGateway(\n",
     "    gateway_uri=\"http://localhost:8000\",  # replace with service URL or host/port of Javelin\n",
     "    route=route_completions,\n",
-    "    model_name=\"text-davinci-003\",\n",
+    "    model_name=\"gpt-3.5-turbo-instruct\",\n",
     ")\n",
     "\n",
     "prompt = PromptTemplate(\"Translate the following English text to French: {text}\")\n",

docs/docs/integrations/llms/llm_caching.ipynb

@@ -21,7 +21,7 @@
     "from langchain.llms import OpenAI\n",
     "\n",
     "# To make the caching really obvious, lets use a slower model.\n",
-    "llm = OpenAI(model_name=\"text-davinci-002\", n=2, best_of=2)"
+    "llm = OpenAI(model_name=\"gpt-3.5-turbo-instruct\", n=2, best_of=2)"
    ]
   },
   {
@@ -1159,7 +1159,7 @@
    "metadata": {},
    "outputs": [
     {
-     "name": "stdin",
+     "name": "stdout",
      "output_type": "stream",
      "text": [
       "ASTRA_DB_API_ENDPOINT =  https://01234567-89ab-cdef-0123-456789abcdef-us-east1.apps.astra.datastax.com\n",
@@ -1358,7 +1358,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "llm = OpenAI(model_name=\"text-davinci-002\", n=2, best_of=2, cache=False)"
+    "llm = OpenAI(model_name=\"gpt-3.5-turbo-instruct\", n=2, best_of=2, cache=False)"
    ]
   },
   {
@@ -1442,8 +1442,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "llm = OpenAI(model_name=\"text-davinci-002\")\n",
-    "no_cache_llm = OpenAI(model_name=\"text-davinci-002\", cache=False)"
+    "llm = OpenAI(model_name=\"gpt-3.5-turbo-instruct\")\n",
+    "no_cache_llm = OpenAI(model_name=\"gpt-3.5-turbo-instruct\", cache=False)"
    ]
   },
   {

docs/docs/integrations/providers/log10.mdx

@@ -63,7 +63,7 @@ llm = ChatAnthropic(model="claude-2", callbacks=[log10_callback], temperature=0.
 llm.predict_messages(messages)
 print(completion)
 
-llm = OpenAI(model_name="text-davinci-003", callbacks=[log10_callback], temperature=0.5)
+llm = OpenAI(model_name="gpt-3.5-turbo-instruct", callbacks=[log10_callback], temperature=0.5)
 completion = llm.predict("You are a ping pong machine.\nPing?\n")
 print(completion)
 ```

docs/docs/integrations/providers/predictionguard.mdx

@@ -88,7 +88,7 @@ os.environ["OPENAI_API_KEY"] = "<your OpenAI api key>"
 # Your Prediction Guard API key. Get one at predictionguard.com
 os.environ["PREDICTIONGUARD_TOKEN"] = "<your Prediction Guard access token>"
 
-pgllm = PredictionGuard(model="OpenAI-text-davinci-003")
+pgllm = PredictionGuard(model="OpenAI-gpt-3.5-turbo-instruct")
 
 template = """Question: {question}
 

docs/docs/integrations/toolkits/openapi.ipynb

@@ -222,7 +222,7 @@
    "source": [
     "import tiktoken\n",
     "\n",
-    "enc = tiktoken.encoding_for_model(\"text-davinci-003\")\n",
+    "enc = tiktoken.encoding_for_model(\"gpt-4\")\n",
     "\n",
     "\n",
     "def count_tokens(s):\n",

docs/docs/integrations/toolkits/openapi_nla.ipynb

@@ -40,9 +40,9 @@
    },
    "outputs": [],
    "source": [
-    "# Select the LLM to use. Here, we use text-davinci-003\n",
+    "# Select the LLM to use. Here, we use gpt-3.5-turbo-instruct\n",
     "llm = OpenAI(\n",
-    "    temperature=0, max_tokens=700\n",
+    "    temperature=0, max_tokens=700, model_name=\"gpt-3.5-turbo-instruct\"\n",
     ")  # You can swap between different core LLM's here."
    ]
   },

docs/docs/integrations/toolkits/powerbi.ipynb

@@ -15,7 +15,7 @@
     "- It relies on authentication with the azure.identity package, which can be installed with `pip install azure-identity`. Alternatively you can create the powerbi dataset with a token as a string without supplying the credentials.\n",
     "- You can also supply a username to impersonate for use with datasets that have RLS enabled. \n",
     "- The toolkit uses a LLM to create the query from the question, the agent uses the LLM for the overall execution.\n",
-    "- Testing was done mostly with a `text-davinci-003` model, codex models did not seem to perform ver well."
+    "- Testing was done mostly with a `gpt-3.5-turbo-instruct` model, codex models did not seem to perform ver well."
    ]
   },
   {

docs/docs/modules/agents/agent_types/react_docstore.ipynb

@@ -36,7 +36,7 @@
     "    ),\n",
     "]\n",
     "\n",
-    "llm = OpenAI(temperature=0, model_name=\"text-davinci-002\")\n",
+    "llm = OpenAI(temperature=0, model_name=\"gpt-3.5-turbo-instruct\")\n",
     "react = initialize_agent(tools, llm, agent=AgentType.REACT_DOCSTORE, verbose=True)"
    ]
   },

docs/docs/modules/agents/how_to/intermediate_steps.ipynb

@@ -36,7 +36,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "llm = OpenAI(temperature=0, model_name=\"text-davinci-002\")\n",
+    "llm = OpenAI(temperature=0, model_name=\"gpt-3.5-turbo-instruct\")\n",
     "tools = load_tools([\"serpapi\", \"llm-math\"], llm=llm)"
    ]
   },

docs/docs/modules/model_io/llms/llm_caching.mdx

@@ -9,7 +9,7 @@ from langchain.globals import set_llm_cache
 from langchain.llms import OpenAI
 
 # To make the caching really obvious, lets use a slower model.
-llm = OpenAI(model_name="text-davinci-002", n=2, best_of=2)
+llm = OpenAI(model_name="gpt-3.5-turbo-instruct", n=2, best_of=2)
 ```
 
 ## In Memory Cache
@@ -110,8 +110,8 @@ As an example, we will load a summarizer map-reduce chain. We will cache results
 
 
 ```python
-llm = OpenAI(model_name="text-davinci-002")
-no_cache_llm = OpenAI(model_name="text-davinci-002", cache=False)
+llm = OpenAI(model_name="gpt-3.5-turbo-instruct")
+no_cache_llm = OpenAI(model_name="gpt-3.5-turbo-instruct", cache=False)
 ```
 
 

docs/docs/modules/model_io/output_parsers/index.ipynb

@@ -55,7 +55,7 @@
     "from langchain.prompts import PromptTemplate\n",
     "from langchain_core.pydantic_v1 import BaseModel, Field, validator\n",
     "\n",
-    "model = OpenAI(model_name=\"text-davinci-003\", temperature=0.0)\n",
+    "model = OpenAI(model_name=\"gpt-3.5-turbo-instruct\", temperature=0.0)\n",
     "\n",
     "\n",
     "# Define your desired data structure.\n",

docs/docs/modules/model_io/output_parsers/pandas_dataframe.ipynb

@@ -34,7 +34,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "model_name = \"text-davinci-003\"\n",
+    "model_name = \"gpt-3.5-turbo-instruct\"\n",
     "temperature = 0.5\n",
     "model = OpenAI(model_name=model_name, temperature=temperature)"
    ]

docs/docs/modules/model_io/output_parsers/pydantic.ipynb

@@ -35,7 +35,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "model_name = \"text-davinci-003\"\n",
+    "model_name = \"gpt-3.5-turbo-instruct\"\n",
     "temperature = 0.0\n",
     "model = OpenAI(model_name=model_name, temperature=temperature)"
    ]

libs/community/langchain_community/embeddings/edenai.py

@@ -20,7 +20,7 @@ class EdenAiEmbeddings(BaseModel, Embeddings):
 
     model: Optional[str] = None
     """
-    model name for above provider (eg: 'text-davinci-003' for openai)
+    model name for above provider (eg: 'gpt-3.5-turbo-instruct' for openai)
     available models are shown on https://docs.edenai.co/ under 'available providers'
     """
 

libs/community/langchain_community/llms/edenai.py

@@ -45,7 +45,7 @@ class EdenAI(LLM):
 
     model: Optional[str] = None
     """
-    model name for above provider (eg: 'text-davinci-003' for openai)
+    model name for above provider (eg: 'gpt-3.5-turbo-instruct' for openai)
     available models are shown on https://docs.edenai.co/ under 'available providers'
     """
 

libs/community/langchain_community/llms/openai.py

@@ -173,7 +173,7 @@ def is_lc_serializable(cls) -> bool:
 
     client: Any = Field(default=None, exclude=True)  #: :meta private:
     async_client: Any = Field(default=None, exclude=True)  #: :meta private:
-    model_name: str = Field(default="text-davinci-003", alias="model")
+    model_name: str = Field(default="gpt-3.5-turbo-instruct", alias="model")
     """Model name to use."""
     temperature: float = 0.7
     """What sampling temperature to use."""
@@ -657,7 +657,7 @@ def modelname_to_contextsize(modelname: str) -> int:
         Example:
             .. code-block:: python
 
-                max_tokens = openai.modelname_to_contextsize("text-davinci-003")
+                max_tokens = openai.modelname_to_contextsize("gpt-3.5-turbo-instruct")
         """
         model_token_mapping = {
             "gpt-4": 8192,
@@ -737,7 +737,7 @@ class OpenAI(BaseOpenAI):
         .. code-block:: python
 
             from langchain_community.llms import OpenAI
-            openai = OpenAI(model_name="text-davinci-003")
+            openai = OpenAI(model_name="gpt-3.5-turbo-instruct")
     """
 
     @classmethod
@@ -763,7 +763,7 @@ class AzureOpenAI(BaseOpenAI):
         .. code-block:: python
 
             from langchain_community.llms import AzureOpenAI
-            openai = AzureOpenAI(model_name="text-davinci-003")
+            openai = AzureOpenAI(model_name="gpt-3.5-turbo-instruct")
     """
 
     azure_endpoint: Union[str, None] = None

libs/community/langchain_community/llms/promptlayer_openai.py

@@ -31,7 +31,7 @@ class PromptLayerOpenAI(OpenAI):
         .. code-block:: python
 
             from langchain_community.llms import PromptLayerOpenAI
-            openai = PromptLayerOpenAI(model_name="text-davinci-003")
+            openai = PromptLayerOpenAI(model_name="gpt-3.5-turbo-instruct")
     """
 
     pl_tags: Optional[List[str]]

libs/community/tests/integration_tests/chat_models/test_konko.py

@@ -163,7 +163,7 @@ def test_konko_additional_args_test() -> None:
         ChatKonko(model_kwargs={"temperature": 0.2})
 
     with pytest.raises(ValueError):
-        ChatKonko(model_kwargs={"model": "text-davinci-003"})
+        ChatKonko(model_kwargs={"model": "gpt-3.5-turbo-instruct"})
 
 
 def test_konko_token_streaming_test() -> None:

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

@@ -261,7 +261,7 @@ def test_chat_openai_extra_kwargs() -> None:
 
     # Test that "model" cannot be specified in kwargs
     with pytest.raises(ValueError):
-        ChatOpenAI(model_kwargs={"model": "text-davinci-003"})
+        ChatOpenAI(model_kwargs={"model": "gpt-3.5-turbo-instruct"})
 
 
 @pytest.mark.scheduled

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

@@ -50,7 +50,7 @@ def test_openai_extra_kwargs() -> None:
 
     # Test that "model" cannot be specified in kwargs
     with pytest.raises(ValueError):
-        OpenAI(model_kwargs={"model": "text-davinci-003"})
+        OpenAI(model_kwargs={"model": "gpt-3.5-turbo-instruct"})
 
 
 def test_openai_llm_output_contains_model_name() -> None:
@@ -286,7 +286,7 @@ def mock_completion() -> dict:
         "id": "cmpl-3evkmQda5Hu7fcZavknQda3SQ",
         "object": "text_completion",
         "created": 1689989000,
-        "model": "text-davinci-003",
+        "model": "gpt-3.5-turbo-instruct",
         "choices": [
             {"text": "Bar Baz", "index": 0, "logprobs": None, "finish_reason": "length"}
         ],

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

@@ -48,7 +48,7 @@ def mock_completion() -> dict:
         "id": "cmpl-3evkmQda5Hu7fcZavknQda3SQ",
         "object": "text_completion",
         "created": 1689989000,
-        "model": "text-davinci-003",
+        "model": "gpt-3.5-turbo-instruct",
         "choices": [
             {"text": "Bar Baz", "index": 0, "logprobs": None, "finish_reason": "length"}
         ],

libs/langchain/tests/integration_tests/chains/test_react.py

@@ -7,7 +7,7 @@
 
 def test_react() -> None:
     """Test functionality on a prompt."""
-    llm = OpenAI(temperature=0, model_name="text-davinci-002")
+    llm = OpenAI(temperature=0, model_name="gpt-3.5-turbo-instruct")
     react = ReActChain(llm=llm, docstore=Wikipedia())
     question = (
         "Author David Chanoff has collaborated with a U.S. Navy admiral "