-
@@ -23,33 +23,35 @@ >[!IMPORTANT] ->This repository utilizes [SourceSage](https://github.com/Sunwood-ai-labs/SourceSage). Approximately 90% of the release notes, README, and commit messages were generated using [SourceSage](https://github.com/Sunwood-ai-labs/SourceSage) and [claude.ai](https://claude.ai/). +>This repository leverages [SourceSage](https://github.com/Sunwood-ai-labs/SourceSage), and approximately 90% of the release notes, README, and commit messages are generated using [SourceSage](https://github.com/Sunwood-ai-labs/SourceSage) + [claude.ai](https://claude.ai/). >[!NOTE] ->AMATERASU is the successor project to [MOA](https://github.com/Sunwood-ai-labs/MOA). It has evolved to run each AI service on an independent EC2 instance using Docker Compose, enabling easier deployment with Terraform. - +>AMATERASU is the successor project to [MOA](https://github.com/Sunwood-ai-labs/MOA). It has evolved to run each AI service on a separate EC2 instance using Docker Compose, enabling easier deployment with Terraform. ## 🚀 Project Overview -AMATERASU is an automation tool for building an LLM (Large Language Model) platform on AWS. While inheriting the functionality of MOA, it offers more flexible scaling and management by running each service on a separate EC2 instance. +AMATERASU is an automation tool for building an LLM (Large Language Model) platform on AWS. While inheriting functionality from MOA, it provides more flexible scaling and management by running each service on a separate EC2 instance. Key Features: - Simple EC2 instance management using Terraform - Independent EC2 instances and Docker Compose environments for each service -- Service-level scaling and operation +- Scalable and manageable at the service level - Secure communication and access control ## ✨ Main Features -- None (at this time) +- Automated AWS infrastructure construction using Terraform +- Containerization and management of each service using Docker Compose +- Integration with multiple LLM models (OpenAI, Anthropic, Gemini, etc.) +- Model management and billing functionality via Langfuse ## 🔧 Usage -Follow the installation instructions and usage guide in this README to set up AMATERASU. +Please follow the installation instructions and usage described in this README to set up AMATERASU. -## 📦 Installation +## 📦 Installation Instructions 1. Clone the repository: ```bash @@ -60,12 +62,12 @@ cd AMATERASU 2. Set environment variables: ```bash cp .env.example .env -# Edit the .env file and set the necessary credentials +# Edit the .env file and set the necessary credentials (LITELLM_MASTER_KEY, LITELLM_SALT_KEY, OPENAI_API_KEY, ANTHROPIC_API_KEY, GEMINI_API_KEY, GEMINI_API_KEY_IRIS, etc.) ``` 3. Initialize and run Terraform: ```bash -cd terraform +cd spellbook/open-webui/terraform terraform init terraform plan terraform apply @@ -80,142 +82,33 @@ ssh -i "C:\Users\makim\.ssh\AMATERASU-terraform-keypair-tokyo-PEM.pem" ubuntu@i- ## 🆕 What's New -v0.2.0 features a revamped architecture, running each AI service in a separate EC2 instance using Docker Compose. This improves scalability and manageability for each service. The English README has been updated, and images have been added to improve the appearance of the release notes. - -The architecture refresh has added an architecture diagram, system requirements, installation instructions, module structure, deployment methods, operational command examples, detailed directory structures for each module, examples of Docker Compose configuration files (`docker-compose.yml`) and environment variable files (`.env`), SSH connection to each module, and scripts for managing services (start, stop, log display) using Docker Compose. For enhanced security, each EC2 instance is protected by a separate security group, and inter-service communication is controlled within the internal VPC network. - - -## 🌐 Module Structure - -Each module runs using Docker Compose on a separate EC2 instance: - -### open-webui Module (EC2 Instance) -``` -📁 open-webui/ -├── 📄 docker-compose.yml # open-webui and ollama configuration -├── 📄 .env # Environment variable settings -└── 📁 config/ # Configuration files -``` - -Example Configuration (docker-compose.yml): -```yaml -version: '3' -services: - ollama: - image: ollama/ollama - ports: - - "11434:11434" - volumes: - - ./data:/root/.ollama - - open-webui: - image: open-webui/open-webui - ports: - - "3000:3000" - environment: - - OLLAMA_URL=http://ollama:11434 -``` - -### litellm Module (EC2 Instance) -``` -📁 litellm/ -├── 📄 docker-compose.yml # litellm service configuration -├── 📄 .env # API key and other environment variables -└── 📁 config/ # LLM configuration files -``` - -### langfuse Module (EC2 Instance) -``` -📁 langfuse/ -├── 📄 docker-compose.yml # langfuse and DB configuration -├── 📄 .env # Environment variable settings -└── 📁 data/ # PostgreSQL data -``` - -## 🔨 Deployment Command Examples - -Deploying specific modules only: -```bash -# Deploy only the open-webui module -terraform apply -target=module.ec2_open_webui - -# Deploy only the litellm module -terraform apply -target=module.ec2_litellm - -# Deploy only the langfuse module -terraform apply -target=module.ec2_langfuse -``` - -## 💻 Module Management Commands - -Connecting to each EC2 instance: -```bash -# SSH connection script -./scripts/connect.sh open-webui -./scripts/connect.sh litellm -./scripts/connect.sh langfuse -``` - -Docker Compose operations: -```bash -# Run within each instance -cd /opt/amaterasu/[module-name] -docker-compose up -d # Start services -docker-compose down # Stop services -docker-compose logs -f # View logs -``` - -## 🔒 Security Configuration - -- Each EC2 instance is protected by a separate security group -- Inter-service communication is controlled within the internal VPC network -- Only the minimum necessary ports are exposed -- Permission management using IAM roles - -## 📚 Directory Structure - -```plaintext -amaterasu/ -├── terraform/ # Terraform code -│ ├── modules/ # Modules for each EC2 instance -│ ├── main.tf # Main configuration -│ └── variables.tf # Variable definitions -├── modules/ # Docker Compose configuration for each service -│ ├── open-webui/ # open-webui related files -│ ├── litellm/ # litellm related files -│ └── langfuse/ # langfuse related files -├── scripts/ # Operational scripts -└── docs/ # Documentation -``` - -## ⚠️ Important Changes - -- Due to the architecture refresh, upgrading from previous versions requires manual migration following the provided steps. Refer to the upgrade instructions for details. - - -## 📦 Upgrade Instructions +### v0.3.0 Updates -1. Stop the existing environment. -2. Build the environment with the new architecture following the instructions in this README. -3. If data migration is necessary, perform the appropriate steps. (Specific steps are not provided.) +- README.md update: Clearly stated the use of SourceSage and Claude.ai, emphasizing important information. +- Added multiple Claude model definitions: Expanded the Claude models available in Langfuse. (`claude-3.5-haiku-20241022`, `claude-3.5-haiku-latest`, `claude-3.5-sonnet-20240620`, `claude-3.5-sonnet-20241022`, `claude-3.5-sonnet-latest`, `claude-3-haiku-20240307`, `claude-3-opus-20240229`, `claude-3-sonnet-20240229`) +- Added environment variable settings for LLaMA model integration: Simplified integration with various LLaMA model providers. (`LITELLM_MASTER_KEY`, `LITELLM_SALT_KEY`, `OPENAI_API_KEY`, `ANTHROPIC_API_KEY`, `GEMINI_API_KEY`, `GEMINI_API_KEY_IRIS`) +- Added SSH connection information to README.md and updated infrastructure description: Added instructions for SSH connection to the EC2 instance and an explanation of the architecture refresh in v0.2.0. +- Updated English README +- Corrected volume mount in docker-compose.yml +- Changed logging library: Changed from the `logging` module to the `loguru` module. ## 📄 License -This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details. +This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details. -## 👏 Acknowledgements +## 👏 Acknowledgments -Thanks to iris-s-coon and Maki. +Thanks to iris-s-coon and Maki for their contributions. -## 🤝 Contributing +## 🤝 Contributions Contributions are welcome! Here's how to get involved: 1. Fork this repository 2. Create a new branch (`git checkout -b feature/amazing-feature`) 3. Commit your changes (`git commit -m 'Add amazing feature'`) -4. Push the branch (`git push origin feature/amazing-feature`) +4. Push your branch (`git push origin feature/amazing-feature`) 5. Create a pull request ## 📧 Support