chore: Cortex Docs should be updated with Docs and API Endpoints

[dan-menlo]

Goal

• Cortex Docs and README should be up-to-date with latest changes
• Goal: by end-Monday, Cortex Docs are good enough to
• Goal: epics start with Docs first, and then once approved coding can start

I would like the full team to understand how the OpenAPI format works, to enable us to handle documentation scalably.

Docs Site

Overall

• Document how to create Redirects for Docusaurus @hiento09
• Are we able to split up OpenAPI file into smaller parts? (e.g. by route)

Get Started

Installation

Owner: @hiento09

• Installation Page (i.e,. how does our installer work)
  • Differences between cortex stable, beta, nightly. Remove repetition from elsewhere
  • Local vs network installer
  • Docker @hiento09 (thank you!)

Quickstart
Basic Usage

Owner: @gabrielle-ong

• Combine Basic Usage and API Server

Architecture

• Cortex Data Folder @hiento09
• cortex.db @vansangpfiev
• .cortexrc @vansangpfiev
  • e.g. What fields are in .cortexrc
• Updater @vansangpfiev
  • e.g. how Updater works

Capabilities

We should talk about capabilities, and link to the appropriate API References.

• Pulling Models @namchuai (thank you!)
  • Model Formats (i.e. our Cortex Model repo format), or Huggingface GGUF repo
• Running Models @nguyenhoangthuan99
  • Should explain runtime params vs. model.yaml
  • Should explain Models use Engines
• Engine Management @namchuai
  • Should give high-level overview of Engine Management, e.g. different types, Versions, Variants
  • llama.cpp (should give instructions on how to install, update, change versions/variants, explain AVX2, AVX-512 etc)
  • Building Engine Extensions (should give instructions on EngineI.h)
• Hardware Awareness @vansangpfiev
• Text Generation
• Embeddings

Guides

• Function Calling @nguyenhoangthuan99
• Structured Outputs @nguyenhoangthuan99

API Reference

Chat

Assignee: @nguyenhoangthuan99

API Reference

• /chat/completions

Pulling Models

Assignee: @namchuai

API Reference

• Pull a Model
• Stop Model Download
• Import Model

Running Models

Assignee: @vansangpfiev

API Reference

• List Models
• Start Models
• Stop Model
• Get Model
• Delete model
• Update model

Engines

Assignee: @namchuai

API Reference

• List available engines
• Get an engine
• Install an engine
• Uninstall an engine

Server

Assignee: @vansangpfiev

Docs

• Basic Usage

API Reference

• Start, Stop
• Update
• Terminate
• Check health

Hardware

Assignee: @vansangpfiev (write Docs before implementing, keep in PR)

Docs

• Hardware Awareness section

API Reference

• Hardware APIs

[gabrielle-ong]

Leftover chores from #1575 - @gabrielle-ong

• CLI commands - remove nightly, beta commands
• Advanced Installation
  • Network vs Local Installer (in Advanced Installation)
  • linux binary links
• Github readme - add Linux binary

[hiento09]

Redirect Instruction:

• Create a file docs/_redirects.
• Add a redirect using the following syntax: source_path destination_path code, for example, /team /about/team 302. This means https://cortex.so/team will redirect to https://cortex.so/about/team.
• File docs/_redirects example

  /home301 / 301
  /home302 / 302
  /querystrings /?query=string 301
  /twitch https://twitch.tv
  /trailing /trailing/ 301
  /notrailing/ /nottrailing 301
  /page/ /page2/#fragment 301
  /blog/* https://blog.my.domain/:splat
  /products/:code/:name /products?code=:code&name=:name
  

Refer to:

• https://github.com/janhq/jan/blob/dev/docs/_redirects
• https://developers.cloudflare.com/pages/configuration/redirects/

cc @janhq/cortex @dan-homebrew @gabrielle-ong

[vansangpfiev]

For Server API endpoints, we only have

• Terminate (Stop)
• Check health
  We don't have Start and Update endpoints. CLI is taking the responsibility for updating.