update tutorials, how-to guides, and high traffic integrations to use init_chat_model

[efriis]

Privileged issue

• I am a LangChain maintainer, or was asked directly by a LangChain maintainer to create an issue here.

Issue Content

These guides (including ones operating off of ChatModelTabs) should use langchain.chat_models.init_chat_model

[rawathemant246]

@efriis i have a question about this

for example

import ChatModelTabs from "@theme/ChatModelTabs";

<ChatModelTabs openaiParams={model="gpt-4"} />

it will be revisied as

from langchain.chat_models import init_chat_model

model = init_chat_model(model= "gpt-4o", model_provider="openai", temperature=0)

like this right?

[efriis]

Not quite! The chat model tabs should be in the same place, but the implementation of the chat model tabs should instantiate the model using init_chat_model instead

[hesreallyhim]

I'm ramping up on the platform and am not opinionated about this issue, but... to me this is confusing (which is not to say it's not correct). When I read the changelog for v3 I saw the transition from:

For example, say that your code is still using the old import from langchain.chat_models import ChatOpenAI:
After the first run, you’ll get: from langchain_community.chat_models import ChatOpenAI After the second run, you’ll get: from langchain_openai import ChatOpenAI

So now it looks like we're going another step to from langchain.chat_models import init_chat_model... but according to the docs, you still have to have the langchain-<provider> package installed(?) (https://python.langchain.com/docs/how_to/chat_models_universal_init/)

I would love to see/understand some more context about this, what is the overarching goal with init_chat_models?

[Sakshamliftin]

Hey! I took a look at the current ChatModelTabs.js implementation and it looks like the code already uses:

from langchain.chat_models import init_chat_model
model = init_chat_model(...)