Refactor AgentJet Swarm Documentation and Architecture

- Deleted outdated swarm blueprint and main documentation files.
- Added new detailed Swarm Design Blueprint and overview documentation in English.
- Updated mkdocs configuration to reflect new documentation structure.
- Introduced a comprehensive blog post on the AgentJet Swarm framework, highlighting its architecture, advantages, and use cases for distributed multi-agent LLM training.

Author: binary-husky

README.md

@@ -18,7 +18,7 @@
 
 
 
-## ✈️ Minimum Example
+## ✈️ Fast Introduction
 
 ### Classic Mode
 
@@ -29,22 +29,32 @@ Let's begin with the simplest example: a math agent with a tool call. This is a
     ```python
     ajet --conf ./tutorial/example_math_agent/math_agent.yaml --backbone='verl'
     ```
+<div align="center">
+<img width="640" alt="image" src="https://serve.gptacademic.cn/publish/shared/Image/classic+swarm+revise.jpg"/>
+</div>
 
 ### Swarm Mode
 
+Let's begin with the simplest AgentJet Swarm example: also a math agent. In this case, you can use any GPU-less laptop to train the model remotely.
+
 1. Start swarm server and begin swarm overwatch: `ajet-swarm start` and `ajet-swarm overwatch --swarm-url=http://localhost:10086`.
-2. From another device (or localhost), run [this script to train](https://github.com/modelscope/AgentJet/blob/main/tutorial/example_math_swarm/math.py):
+2. From your laptop (or swarm server localhost), run [this script to train](https://github.com/modelscope/AgentJet/blob/main/tutorial/example_math_swarm/math.py) to start training:
     ```python
     AJET_SWARM_URL="http://swarm-server-ip:10086" python ./tutorial/example_math_swarm/math.py
     ```
+<div align="center">
+<img width="600" alt="image" src="https://serve.gptacademic.cn/publish/shared/Image/swarm-server.gif"/>
+</div>
+
 
 ## ✈️ Features
 
 We aim to build a easy-to-learn Agent tuner that unlock more possibilities for agent developers:
 
 - **Easy and Friendly**. AgentJet helps you tune models behind your agent workflows easily, optimizing your agents for top performance with minimal effort.
 - **Rich Tutorial Library**. AgentJet provides a rich library of [examples](https://github.com/modelscope/AgentJet/tree/main/tutorial) as tutorials.
-- **Efficient and Scalable**. AgentJet uses [verl] as the default backbone (`--backbone=verl`). However, we also support [trinity](https://github.com/modelscope/Trinity-RFT/) as alternative backbone, accelerating your tuning process via fully asynchronous RFT.
+- **Swarm Training**. [This unique feature](https://modelscope.github.io/AgentJet/en/swarm_intro_blog_english/) of AgentJet opens many possibilities: deploying distributed & self-healing rollout workers, **non-shared-parameter multi-agent** training, **multi-runtime & multi-task cocktail** training. And just like Tinker, you can use AgentJet Swarm to train **models even on **GPU-less laptop(s)**.
+- **Efficient and Scalable**. AgentJet uses [verl] as the default backbone (`--backbone=verl`). However, we also support trinity as alternative backbone, accelerating your tuning process via fully asynchronous RFT.
 - **Flexible and Fast**. AgentJet supports [multi-agent workflows](https://modelscope.github.io/AgentJet/en/workflow/) and adopts a context merging technique, accelerating training by 1.5x to 10x when the workflow involves multi-turn (or multi-agent) conversations.
 - **Reliability and Reproducibility**. Our team keeps track of framework performance across multiple [tasks + major-git-version + training-backbones](https://benchmark.agentjet.top/) (under construction, still gathering data, coming soon).
 
@@ -62,13 +72,6 @@ For advanced researchers, AgentJet also provides high-resolution logging and deb
 
 - **Click here to read the** [**installation guide**](https://modelscope.github.io/AgentJet/en/installation/).
 
-#### Run Training
-
-- You can start training your first agent with a single command using a pre-configured YAML file. Take the [Math agent](https://modelscope.github.io/AgentJet/en/example_math_agent/) as an example:
-
-  ```bash
-  ajet --conf tutorial/example_math_agent/math_agent.yaml
-  ```
 
 #### Example Library
 
@@ -111,6 +114,7 @@ The internal system orchestrates several specialized modules to handle the compl
 * **Task Runner**: Executes the Agent workflow and calculates rewards.
 * **Model Tuner**: Forwards inference requests from the workflow to the LLM engine.
 * **Context Tracker**: Monitors LLM calls and automatically merges shared-history timelines to improve training efficiency by **1.5x to 10x**.
+* **Swarm Server**: A data interchange center that accept OpenAI-like requests and engine instructions, activated only in AgentJet Swarm mode.
 
 
 
@@ -128,14 +132,11 @@ AgentJet is a constantly evolving project. We are planning to add the following
 
 | Category | Feature | Status |
 | :--- | :--- | :--- |
-| **Examples** | Covering LangGraph and AutoGen frameworks | Done & Verifying |
 | **Examples** | Add LoRA training examples | Todo |
-| **Infra** | Cross-process Tuner wrapper to pass though process forking | Done & Verifying |
 | **Infra** | Optimize configurations for long-context adaptation on smaller GPUs | In Progress |
-| **Capability** | Prompt tuning | In Progress |
 | **Capability** | Multi-modal training support | Todo |
 | **Capability** | MARL Credit assignment | Todo |
-| **Capability** | Training dataset generation from few-shot samples | Done & Verifying |
+| **Capability** | Training dataset generation from few-shot samples | Todo |
 
 
 ## ✈️ Citation
@@ -158,8 +159,9 @@ If you use AgentJet in your research, please cite:
 
 ---
 <div align="center">
+This project is under active development, we need your help to make it shine! <br/>
 
-[⭐ Star Us](https://github.com/modelscope/AgentJet) · [Report Bug](https://github.com/modelscope/AgentJet/issues) · [Request Feature](https://github.com/modelscope/AgentJet/issues)
+[⭐ Star Us](https://github.com/modelscope/AgentJet) · [✈️ Report Bug](https://github.com/modelscope/AgentJet/issues) · [✈️ Request Feature](https://github.com/modelscope/AgentJet/issues)
 </div>
 
 

agentjet-swarm-blueprint.md docs/en/agentjet-swarm-blueprint.mdagentjet-swarm-blueprint.md renamed to docs/en/agentjet-swarm-blueprint.md

@@ -31,7 +31,7 @@ Unlike traditional setups where the user code must run inside the training clust
 | **架构模式** | 托管服务 + 单点客户端 API | **✅ 服务器和客户端都可按需拓展** |
 | **多客户端共同参与训练** | ❌ 不支持 | **✅ 支持** |
 | **远程算力部署方式** | Thinking Machines Lab 公司提供定价 | **✅ 自建 GPU 服务器端 或 使用阿里云灵骏** |
-| **训练方式** | ❌ LoRA 微调 | **✅ 全量 LLM 模型训练** |
+| **训练方式** |  仅限 LoRA 微调 | **✅ 全量 LLM 模型训练** |
 | **支持的模型** | ❌ 少部分 LLM 模型 | **✅ 大多数新旧 LLM 模型** |
 | **最大模型规模** | Llama 70B、Qwen 235B | **✅ 取决于用户 GPU 集群配置** |
 | **通信协议** | 专有 API | **✅ 专有API + OpenAI兼容API** |