Merge pull request #311 from janhq/0.6.0-docs

Update docs for the newest version

src/pages/docs/_meta.json

@@ -19,10 +19,10 @@
   },
   "models": "Models",
   "tools": "Tools",
+  "assistants": "Assistants",
   "threads": "Threads",
   "settings": "Settings",
   "shortcuts": "Keyboard Shortcuts",
-  "local-api": "",
   "inference-engines": {
     "title": "MODEL PROVIDER",
     "type": "separator"

src/pages/docs/assistants.mdx

@@ -0,0 +1,34 @@
+---
+title: Assistants
+description: A step-by-step guide on customizing your assistant.
+keywords:
+  [
+    Jan,
+    Customizable Intelligence, LLM,
+    local AI,
+    privacy focus,
+    free and open source,
+    private and offline,
+    conversational AI,
+    no-subscription fee,
+    large language models,
+    manage assistants,
+    assistants,
+  ]
+---
+
+import { Callout, Steps } from 'nextra/components'
+
+# Assistants
+This guide explains how to set the Assistant instructions in the Jan application.
+
+## Applied the Instructions to All Threads
+To apply the instructions to all the new threads, follow these steps:
+1. Select a **Thread**.
+2. Click the **Assistant** tab.
+3. Toggle the **slider** to ensure these instructions are applied to all new threads. (Activate the **Experimental Mode** feature to enable this option.)
+<br/>
+
+![Assistant Slider](./_assets/assistant-slider.png)
+
+<br/>

src/pages/docs/built-in/_meta.json

@@ -0,0 +1,10 @@
+{
+  "llama-cpp": {
+    "title": "llama.cpp",
+    "href": "/docs/built-in/llama-cpp"
+  },
+  "tensorrt-llm": {
+    "title": "TensorRT-LLM",
+    "href": "/docs/built-in/tensorrt-llm"
+  }
+}

src/pages/docs/built-in/llama-cpp.mdx

@@ -1,6 +1,6 @@
 ---
 title: llama.cpp
-description: A step-by-step guide on how to customize the llama.cpp extension.
+description: A step-by-step guide on how to customize the llama.cpp engine.
 keywords:
   [
     Jan,
@@ -13,7 +13,7 @@ keywords:
     no-subscription fee,
     large language models,
     Llama CPP integration,
-    llama.cpp Extension,
+    llama.cpp Engine,
     Intel CPU,
     AMD CPU,
     NVIDIA GPU,
@@ -30,59 +30,28 @@ import { Callout, Steps } from 'nextra/components'
 
 ## Overview
 
-Jan has a default [C++ inference server](https://github.com/janhq/nitro) built on top of [llama.cpp](https://github.com/ggerganov/llama.cpp). This server provides an OpenAI-compatible API, queues, scaling, and additional features on top of the wide capabilities of `llama.cpp`.
+Jan has a default [C++ inference server](https://github.com/janhq/cortex) built on top of [llama.cpp](https://github.com/ggerganov/llama.cpp). This server provides an OpenAI-compatible API, queues, scaling, and additional features on top of the wide capabilities of `llama.cpp`.
 
-## llama.cpp Extension
+## llama.cpp Engine
 
-This guide shows you how to modify your engine's behavior by adjusting its settings in the `model.json` file.
+This guide shows you how to initialize the `llama.cpp` to download and install the required dependencies to start chatting with a model using the `llama.cpp` engine.
 
 ## Prerequisites
-<Tabs items={['Mac', 'Windows', 'Linux']}>
-    <Tabs.Tab >
-      <Tabs items={['Mac Intel', 'Mac Silicon']}>
-        <Tabs.Tab >
-          - Make sure you're using an Intel-based Mac. For a complete list of supported Intel CPUs, please see [here](https://en.wikipedia.org/wiki/MacBook_Pro_(Intel-based)).
-          - For Mac Intel, it is recommended to utilize smaller models.
-          <Callout type="info">
-            This uses CPU by default, and no acceleration option is available. 
-          </Callout>
-
-        </Tabs.Tab>
-
-        <Tabs.Tab >
-          - Make sure you're using a Mac Silicon. For a complete list of supported Apple Silicon CPUs, please see [here](https://en.wikipedia.org/wiki/Apple_Silicon).
-          - Using an adequate model size based on your hardware is recommended for Mac Silicon.
-          <Callout type="info">
-            This can use Apple GPU with Metal by default for acceleration. Apple ANE is not supported yet.
-          </Callout>
-        </Tabs.Tab>
-
-      </Tabs>
-    </Tabs.Tab>
-    <Tabs.Tab >
-      - Ensure that you have **Windows with x86_64** architecture.
-      - Select a model size suited to your Windows hardware by looking for the `Recommended RAM` tag in the Hub.
-
-      <Callout type="info">
-        - It will use CPU by default if you do not have any GPU/ NPU.
-        - Ensure that you have also installed the correct CPU instruction.
-      </Callout>
-
-
-    </Tabs.Tab>
-    <Tabs.Tab >
-        - Ensure that you have **Linux with x86_64** architecture.
-        - Select a model size suited to your Linux hardware based on the `Recommended` tag for `RAM` in the Hub.
-
-      <Callout type="info">
-        - It will use CPU by default if you do not have any GPU/ NPU.
-        - Ensure that you have also installed the correct CPU instruction.
-      </Callout>
-
-    </Tabs.Tab>
-</Tabs>
+- Mac Intel:
+  - Make sure you're using an Intel-based Mac. For a complete list of supported Intel CPUs, please see [here](https://en.wikipedia.org/wiki/MacBook_Pro_(Intel-based)).
+  - For Mac Intel, it is recommended to utilize smaller models.
+- Mac Sillicon:
+  - Make sure you're using a Mac Silicon. For a complete list of supported Apple Silicon CPUs, please see [here](https://en.wikipedia.org/wiki/Apple_Silicon).
+  - Using an adequate model size based on your hardware is recommended for Mac Silicon.
+<Callout type="info">
+  This can use Apple GPU with Metal by default for acceleration. Apple ANE is not supported yet.
+</Callout>
+- Windows:
+  - Ensure that you have **Windows with x86_64** architecture.
+- Linux:
+  - Ensure that you have **Linux with x86_64** architecture.
 
-### GPU Acceleration Options
+#### GPU Acceleration Options
 Enable the GPU acceleration option within the Jan application by following the [Installation Setup](/docs/desktop-installation) guide.
 ## Step-by-step Guide
 <Steps>

src/pages/docs/built-in/tensorrt-llm.mdx

@@ -1,6 +1,6 @@
 ---
 title: TensorRT-LLM
-description: A step-by-step guide on customizing the TensorRT-LLM extension.
+description: A step-by-step guide on customizing the TensorRT-LLM engine.
 keywords:
   [
     Jan,
@@ -12,10 +12,10 @@ keywords:
     conversational AI,
     no-subscription fee,
     large language models,
-    TensorRT-LLM Extension,
+    TensorRT-LLM Engine,
     TensorRT,
     tensorRT,
-    extension,
+    engine,
   ]
 ---
 
@@ -25,29 +25,20 @@ import { Callout, Steps } from 'nextra/components'
 
 ## Overview
 
-This guide walks you through installing Jan's official [TensorRT-LLM Extension](https://github.com/janhq/nitro-tensorrt-llm). This extension uses [Nitro-TensorRT-LLM](https://github.com/janhq/nitro-tensorrt-llm) as the AI engine instead of the default [Nitro-Llama-CPP](https://github.com/janhq/nitro).  It includes an efficient C++ server that executes the [TRT-LLM C++ runtime](https://nvidia.github.io/TensorRT-LLM/gpt_runtime.html) natively. It also includes features and performance improvements like OpenAI compatibility, tokenizer improvements, and queues.
+This guide walks you through installing Jan's official [TensorRT-LLM Engine](https://github.com/janhq/nitro-tensorrt-llm). This engine uses [Cortex-TensorRT-LLM](https://github.com/janhq/cortex.tensorrt-llm) as the AI engine instead of the default [Cortex-Llama-CPP](https://github.com/janhq/cortex).  It includes an efficient C++ server that executes the [TRT-LLM C++ runtime](https://nvidia.github.io/TensorRT-LLM/gpt_runtime.html) natively. It also includes features and performance improvements like OpenAI compatibility, tokenizer improvements, and queues.
 
 <Callout type='warning' emoji="">
-  - This feature is only available for Windows users. Linux is coming soon.
-
-  - Additionally, we only prebuilt a few demo models. You can always build your desired models directly on your machine. For more information, please see [here](#build-your-own-tensorrt-models).
+ This feature is only available for Windows users. Linux is coming soon.
 </Callout>
 
 ### Pre-requisites
 
-- A Windows PC
-- Nvidia GPU(s): Ada or Ampere series (i.e. RTX 4000s & 3000s). More will be supported soon.
-- 3GB+ of disk space to download TRT-LLM artifacts and a Nitro binary
-- Jan v0.4.9+ or Jan v0.4.8-321+ (nightly)
-- Nvidia Driver v535+ (For installation guide, please see [here](/docs/troubleshooting#1-ensure-gpu-mode-requirements))
-- CUDA Toolkit v12.2+ (For installation guide, please see [here](/docs/troubleshooting#1-ensure-gpu-mode-requirements))
+- A **Windows** PC.
+- **Nvidia GPU(s)**: Ada or Ampere series (i.e. RTX 4000s & 3000s). More will be supported soon.
+- Sufficient disk space for the TensorRT-LLM models and data files (space requirements vary depending on the model size).
 
-<Callout type='warning'>
-  If you are using our nightly builds, you may have to reinstall the TensorRT-LLM extension each time you update the app. We're working on better extension lifecycles - stay tuned.
-</Callout>
 
 <Steps>
-
 ### Step 1: Install TensorRT-Extension
 
 1. Click the **Gear Icon (⚙️)** on the bottom left of your screen.
@@ -65,7 +56,7 @@ This guide walks you through installing Jan's official [TensorRT-LLM Extension](
 3. Check that files are correctly downloaded.
 
 ```bash
-ls ~/jan/extensions/@janhq/tensorrt-llm-extension/dist/bin
+ls ~/jan/data/extensions/@janhq/tensorrt-llm-extension/dist/bin
 # Your Extension Folder should now include `nitro.exe`, among other artifacts needed to run TRT-LLM
 ```
 
@@ -97,17 +88,4 @@ We offer a handful of precompiled models for Ampere and Ada cards that you can i
 <br/>
 ![Specific Conversation](../_assets/model-parameters.png)
 
-</Steps>
-
-## Build your own TensorRT Engine
-To create custom TensorRT engines, you can follow the step-by-step guide provided in the [NVIDIA documentation](https://nvidia.github.io/TensorRT-LLM/quick-start-guide.html#compile-the-model-into-a-tensorrt-engine).
-
-When compiling and running these models, please adhere to the following compatibility guidelines:
-- **GPU Architectures**: Models are specifically compiled for certain GPU architectures, such as Ada. Ensure that your model is compatible with the architecture of the GPU on which it will run.
-- **TensorRT-LLM Release**: Models need to be compiled and run on the same version of the TensorRT-LLM. For example, a model compiled with version 0.9.0 must be run on version 0.9.0.
-- **Operating System Compatibility**: As of version 0.9.0, models are designed to be cross-OS compatible. However, this feature is still under evaluation and might exhibit instability.
-- **GPU Topology**: It is crucial to understand your system's GPU topology, especially when dealing with multiple GPUs. This can be determined by the number of engine files in use.
-
-<Callout type='info'>
-Ensure these parameters are aligned correctly to avoid runtime issues and fully leverage TensorRT engines' capabilities.
-</Callout>
+</Steps>

src/pages/docs/data-folder.mdx

@@ -25,7 +25,32 @@ import { Callout, Steps } from 'nextra/components'
 # Jan Data Folder
 Jan stores your data locally in your own filesystem in a universal file format (JSON). We build for privacy by default and do not collect or sell your data.
 
-This guide helps you understand where and how this data is stored. We'll also show you how to delete or move the data folder location.
+This guide helps you understand where and how this data is stored.
+
+## Open the Data Folder
+
+To open the Jan data folder from the app:
+1. Click the System monitor button on your Jan app.
+2. Click the App Log button.
+3. This redirects you to the Jan data folder.
+
+```bash
+# Windows
+~/AppData/Roaming/Jan/data
+
+# Mac
+~/Library/Application\ Support/Jan/data
+
+# Linux
+## Custom installation directory
+$XDG_CONFIG_HOME = /home/username/custom_config
+
+or
+
+## Default installation directory
+~/.config/Jan/data
+
+```
 
 ## Folder Structure
 Jan app data folder should have the following folder structure:
@@ -36,8 +61,6 @@ Jan is stored in the root `~/jan` by default.
 /assistants
   /jan
     assistant.json
-  /shakespeare
-    assistant.json
 /extensions
   extensions.json
   /@janhq
@@ -47,7 +70,8 @@ Jan is stored in the root `~/jan` by default.
   /app.txt
 /models
   /model_A
-    model.json
+    model.yaml
+  model_A.yaml
 /settings
   settings.json
   /@janhq
@@ -58,6 +82,11 @@ Jan is stored in the root `~/jan` by default.
   /joi-dark
   /joi-light
   /night-blue
+/themes
+  /dark-dimmed
+  /joi-dark
+  /joi-light
+  /night-blue
 /threads
   /jan_thread_A
    messages.jsonl
@@ -100,7 +129,6 @@ Each parameter in the file is defined as follows:
 | description | Describes the assistant’s capabilities and intended role. |
 | model | Defines accessible models, with "*" representing access to all models. |
 | instructions | Specifies queries and commands to tailor Jan's responses for improved interaction effectiveness. |
-- **Custom Assistant Example**: The `/assistants/shakespeare/` shows a custom setup, also with its own `assistant.json`.
 
 ### `extensions/`
 
@@ -126,6 +154,7 @@ General settings for the application are stored here, separate from individual a
 - **General Settings**: The `settings.json` in the `/settings/` directory holds application-wide settings.
 - **Extension-specific Settings**: Additional settings for extensions are stored in respective subdirectories under `/settings/@janhq/`.
 
+
 ### `themes/`
 
 The `themes` directory contains different visual themes for the application, allowing customization of the user interface.
@@ -134,6 +163,7 @@ The `themes` directory contains different visual themes for the application, all
 
 Threads history is kept in this directory. Each session or thread is stored in a way that makes it easy to review past interactions. Each thread is stored in its subdirectory, such as `/threads/jan_unixstamp/`, with files like `messages.jsonl` and `thread.json` detailing the thread settings.
 
+
 ## Open the Data Folder
 
 To open the Jan data folder, follow the steps in the [Settings](/docs/settings#access-the-jan-data-folder) guide.