[Documentation] - Skeleton for documentation

[clun]

Context

After chatting with the owners of the framework, we realized it was a good moment to start providing more documentation.
As I have set up documentation using MKDocs + Materials previously I wanted to propose a template to start from.
So here it goes.

If you want to see immediately what it looks like

https://clun.github.io/langchain4j-docs

Where is the code ?

https://github.com/clun/langchain4j-docs

How can I run it locally ?

• You need to install python.

• Clone the project

git clone https://github.com/clun/langchain4j-docs.git
cd langchain4j-docs

• Install the dependencies

python3 -m pip install --upgrade pip     # install pip
python3 -m pip install mkdocs            # install mkdocs 
python3 -m pip install mkdocs-material   # install material theme
python3 -m pip install mkdocs-git-revision-date-plugin # install git revision date
python3 -m pip install mkdocs-video # install git revision date
python3 -m pip install mkdocs-video # install git revision date

• Start locally and see site available at http://localhost:8000

mkdocs serve

• Any modification in the files is immediately hot reloaded and you see the updates

How to we deploy ?

That is the neat part. With a github action here on any commit or pull request accepted the webiste is regenerated in the gh-pages !

How do you get the code in the organization ?

• If you are happy with the skeleton I transfer the ownership to this organization

• it would be nice to add me CLUN to keep contributing.

[langchain4j]

Hi @clun thank you so much! I will play with it a bit and will start adding content.

[clun]

Of course.

• I requested the ownership transfer to this org.

• the page cheat.md https://clun.github.io/langchain4j-docs/cheat/ help to know what is doable in the markdown.

[LizeRaes]

@clun Thank you so much for starting this website, looking great! We're happy to extend and complete it, getting @Martin7-1 on board here too

After a close look, I see following action items:

• on the 'Getting Started' page ( https://clun.github.io/langchain4j-docs/pages/documentation/introduction/ )
  I suggest adding more examples to 'Getting Started', next to HelloWorld, suggestions:
  * Chat
  * Streaming
  * Classifier
  * RAG
  * Using tools / calling functions
  * Image generation
• I would move modules to it's own tab, and add (description, example, tutorial)
  * Image Creation
  * Tools / function calling
  * AI Services
  * (to be removed?) Agents (I don't think they are supported yet)
• under 'Integrations' ( https://clun.github.io/langchain4j-docs/pages/integrations/ ), include the lists of supported
  * model integrations (language models + image models)
  * document parsers
  * vector/embedding stores
  * embedding models
  * code execution APIs
  * frameworks (GraalVM?, Quarkus, Spring Boot, Azure OpenAI SDK, anything missing?)
  (optionally add examples or tutorials for each)
• change 'Sample Codes' to 'Code Examples'? https://clun.github.io/langchain4j-docs/pages/samples/
  Add more in-depth examples (invite contributors to put their code there, we can curate), also link the examples from https://github.com/langchain4j/langchain4j-examples and optionally turn it into tutorials
• There are a couple of blocks on the start page, we could make them clickable and link to doc/examples
  • Assistant and ChatBot
  • Vector Stores
  • Embeddings Computation
  • Chains
  • (to be added) AI Services
  • (to be added) Image Generation
• ensure the website stays up-to-date: make it update relevant fields/pages automatically with every new release, and keep track of the things that have to be manually reviewed/updated every release
• if we write tutorials, we could publish some on eg. Baeldung, Medium (langchain4j channel: https://medium.com/@langchain4j ), ... so people do find them when searching for things like
  • 'how to connect to gpt4/LLMs with java'
  • 'how to use AI in java/in the backend'
  • 'generate AI images via java'

Note that quite some examples already exist on https://github.com/langchain4j/langchain4j-examples

I'll get started on listing the integrations.

[langchain4j]

Hi @clun, could you please update us on the status of the transfer? Mabe it will be simpler to just copy-paste it? Thanks

[LizeRaes]

Closing because covered in #429