aws-emd 0.7.0

Easy Model Deployer - Simple, Efficient, and Easy-to-Integrate

---

Latest News 🔥

• [2025/03] We officially released EMD!

---

About

EMD (Easy Model Deployer) is a lightweight tool designed to simplify model deployment. Built for developers who need reliable and scalable model serving without complex setup.

Key Features

• One-click deployment of models to the cloud (Amazon SageMaker, Amazon ECS, Amazon EC2)
• Diverse model types (LLMs, VLMs, Embeddings, Vision, etc.)
• Rich inference engine (vLLM, TGI, Lmdeploy, etc.)
• Different instance types (CPU/GPU/AWS Inferentia)
• Convenient integration (OpenAI Compatible API, LangChain client, etc.)

Notes

• OpenAI Compatible API is supported only for Amazon ECS and Amazon EC2 deployment.

Table of Contents

• Architecture
• Installation
• Usage
• Documentation
• Contributing

Architecture

Deploy models to the cloud with EMD will use the following components in Amazon Web Services:

Getting Started

Installation

Install EMD with pip, currently only support for Python 3.9 and above:

pip install https://github.com/aws-samples/easy-model-deployer/releases/download/main/emd-0.6.0-py3-none-any.whl

Visit our documentation to learn more.

Usage

Choose your default aws profile.

emd config set-default-profile-name

Notes: If you don't set aws profile, it will use the default profile in your env (suitable for Temporary Credentials). Whenever you want to switch deployment accounts, run emd config set-default-profile-name

Bootstrap emd stack

emd bootstrap

Notes: This is going to set up the necessary resources for model deployment. Whenever you change EMD version, run this command again.

Quickly see what models are supported by emd list-supported-models. This command will output all information related to deployment. The following command is recommended to just check the model type. (Plese check Supported Models for complete information.)

emd list-supported-models | jq -r '.[] | "\(.model_id)\t\(.model_type)"' | column -t -s $'\t' | sort

Choose deployment parameters interactively by emd deploy or deploy with one command

emd deploy --model-id DeepSeek-R1-Distill-Qwen-1.5B --instance-type g5.8xlarge --engine-type vllm --framework-type fastapi --service-type sagemaker --extra-params {} --skip-confirm

Notes: Get complete parameters by emd deploy --help and find the values of the required parameters here When you see "Waiting for model: ...", it means the deployment task has started, you can quit the current task by ctrl+c.

Check deployment status.

emd status

Notes: EMD allows to launch multiple deployment tasks at the same time.

Quick functional verfication or check our documentation for integration examples.

emd invoke DeepSeek-R1-Distill-Qwen-1.5B

Notes: Find ModelId in the output of emd status.

Delete the deployed model

emd destroy DeepSeek-R1-Distill-Qwen-1.5B

Notes: Find ModelId in the output of emd status.

Documentation

For advanced configurations and detailed guides, visit our documentation site.

Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.