Today, we are thrilled to announce that DeepSeek R1 distilled Llama and Qwen designs are available through Amazon Bedrock Marketplace and Amazon SageMaker JumpStart. With this launch, you can now release DeepSeek AI's first-generation frontier design, DeepSeek-R1, in addition to the distilled versions ranging from 1.5 to 70 billion specifications to construct, experiment, and responsibly scale your generative AI concepts on AWS.
In this post, we show how to begin with DeepSeek-R1 on Amazon Bedrock Marketplace and SageMaker JumpStart. You can follow similar steps to deploy the distilled versions of the designs too.
Overview of DeepSeek-R1
DeepSeek-R1 is a large language model (LLM) established by DeepSeek AI that uses support learning to improve thinking capabilities through a multi-stage training procedure from a DeepSeek-V3-Base structure. A key identifying feature is its support knowing (RL) step, which was used to refine the model's actions beyond the basic pre-training and fine-tuning procedure. By integrating RL, DeepSeek-R1 can adjust better to user feedback and goals, ultimately enhancing both importance and clarity. In addition, DeepSeek-R1 employs a chain-of-thought (CoT) approach, suggesting it's geared up to break down complex inquiries and factor through them in a detailed manner. This guided reasoning process permits the design to produce more accurate, transparent, and detailed answers. This design combines RL-based fine-tuning with CoT capabilities, aiming to create structured responses while focusing on interpretability and user interaction. With its comprehensive capabilities DeepSeek-R1 has recorded the industry's attention as a flexible text-generation model that can be integrated into numerous workflows such as agents, sensible reasoning and information analysis jobs.
DeepSeek-R1 utilizes a Mixture of Experts (MoE) architecture and is 671 billion specifications in size. The MoE architecture allows activation of 37 billion criteria, enabling effective inference by routing queries to the most pertinent expert "clusters." This approach permits the model to specialize in different problem domains while maintaining overall effectiveness. DeepSeek-R1 needs at least 800 GB of HBM memory in FP8 format for reasoning. In this post, we will utilize an ml.p5e.48 xlarge instance to release the design. ml.p5e.48 xlarge features 8 Nvidia H200 GPUs supplying 1128 GB of GPU memory.
DeepSeek-R1 distilled models bring the reasoning capabilities of the main R1 design to more efficient architectures based on popular open models like Qwen (1.5 B, 7B, 14B, and 32B) and Llama (8B and 70B). Distillation refers to a process of training smaller sized, more efficient models to imitate the habits and thinking patterns of the larger DeepSeek-R1 design, using it as a teacher design.
You can deploy DeepSeek-R1 design either through SageMaker JumpStart or Bedrock Marketplace. Because DeepSeek-R1 is an emerging design, we suggest deploying this design with guardrails in location. In this blog site, we will utilize Amazon Bedrock Guardrails to present safeguards, prevent harmful material, and examine designs against key security requirements. At the time of writing this blog, for DeepSeek-R1 deployments on SageMaker JumpStart and Bedrock Marketplace, Bedrock Guardrails supports only the ApplyGuardrail API. You can develop multiple guardrails tailored to various use cases and apply them to the DeepSeek-R1 model, improving user experiences and standardizing security controls throughout your generative AI applications.
Prerequisites
To release the DeepSeek-R1 model, you need access to an ml.p5e instance. To inspect if you have quotas for P5e, open the Service Quotas console and under AWS Services, select Amazon SageMaker, and confirm you're using ml.p5e.48 xlarge for endpoint use. Make certain that you have at least one ml.P5e.48 xlarge circumstances in the AWS Region you are deploying. To request a limit increase, produce a limitation boost request and reach out to your account team.
Because you will be deploying this design with Amazon Bedrock Guardrails, make certain you have the appropriate AWS Identity and Gain Access To Management (IAM) approvals to use Amazon Bedrock Guardrails. For guidelines, see Set up authorizations to use guardrails for material filtering.
Implementing guardrails with the ApplyGuardrail API
Amazon Bedrock Guardrails permits you to introduce safeguards, avoid damaging content, and examine models against essential security requirements. You can implement security procedures for the DeepSeek-R1 model utilizing the Amazon Bedrock ApplyGuardrail API. This allows you to use guardrails to examine user inputs and design actions deployed on Amazon Bedrock Marketplace and SageMaker JumpStart. You can produce a guardrail utilizing the Amazon Bedrock console or the API. For the example code to create the guardrail, see the GitHub repo.
The general circulation involves the following actions: First, the system gets an input for the design. This input is then processed through the ApplyGuardrail API. If the input passes the guardrail check, it's sent to the design for reasoning. After getting the design's output, another guardrail check is applied. If the output passes this final check, it's returned as the last result. However, if either the input or output is intervened by the guardrail, a message is returned showing the nature of the intervention and whether it took place at the input or output stage. The examples showcased in the following sections show inference using this API.
Deploy DeepSeek-R1 in Amazon Bedrock Marketplace
Amazon Bedrock Marketplace gives you access to over 100 popular, emerging, and specialized foundation models (FMs) through Amazon Bedrock. To gain access to DeepSeek-R1 in Amazon Bedrock, complete the following actions:
1. On the Amazon Bedrock console, pick Model brochure under Foundation models in the navigation pane.
At the time of composing this post, you can utilize the InvokeModel API to conjure up the design. It does not support Converse APIs and other Amazon Bedrock tooling.
2. Filter for DeepSeek as a company and choose the DeepSeek-R1 design.
The model detail page supplies vital details about the design's abilities, pricing structure, and execution standards. You can find detailed usage instructions, including sample API calls and code bits for combination. The design supports different text generation jobs, consisting of material production, code generation, and question answering, utilizing its support finding out optimization and CoT reasoning capabilities.
The page likewise consists of release options and licensing details to assist you start with DeepSeek-R1 in your applications.
3. To start utilizing DeepSeek-R1, pick Deploy.
You will be prompted to set up the release details for DeepSeek-R1. The design ID will be pre-populated.
4. For Endpoint name, go into an endpoint name (in between 1-50 alphanumeric characters).
5. For Number of circumstances, enter a number of circumstances (between 1-100).
6. For Instance type, choose your circumstances type. For optimum performance with DeepSeek-R1, a GPU-based circumstances type like ml.p5e.48 xlarge is advised.
Optionally, you can set up advanced security and facilities settings, including virtual private cloud (VPC) networking, service role approvals, and encryption settings. For a lot of use cases, the default settings will work well. However, for production deployments, you might desire to review these settings to line up with your organization's security and compliance requirements.
7. Choose Deploy to begin using the design.
When the deployment is complete, you can evaluate DeepSeek-R1's abilities straight in the Amazon Bedrock play area.
8. Choose Open in playground to access an interactive interface where you can explore various prompts and change design criteria like temperature and maximum length.
When using R1 with Bedrock's InvokeModel and Playground Console, utilize DeepSeek's chat template for optimal results. For example, content for inference.
This is an exceptional method to check out the design's thinking and text generation capabilities before integrating it into your applications. The playground supplies immediate feedback, helping you understand how the model reacts to various inputs and letting you tweak your triggers for optimum outcomes.
You can quickly evaluate the model in the playground through the UI. However, to invoke the deployed design programmatically with any Amazon Bedrock APIs, you need to get the endpoint ARN.
Run inference using guardrails with the deployed DeepSeek-R1 endpoint
The following code example shows how to carry out reasoning using a released DeepSeek-R1 model through Amazon Bedrock utilizing the invoke_model and ApplyGuardrail API. You can develop a guardrail utilizing the Amazon Bedrock console or the API. For the example code to create the guardrail, see the GitHub repo. After you have produced the guardrail, utilize the following code to implement guardrails. The script initializes the bedrock_runtime customer, configures reasoning specifications, and sends a demand to produce text based on a user timely.
Deploy DeepSeek-R1 with SageMaker JumpStart
SageMaker JumpStart is an artificial intelligence (ML) center with FMs, integrated algorithms, and prebuilt ML options that you can release with just a couple of clicks. With SageMaker JumpStart, you can tailor pre-trained models to your use case, with your information, and deploy them into production utilizing either the UI or SDK.
Deploying DeepSeek-R1 design through SageMaker JumpStart offers two practical approaches: using the user-friendly SageMaker JumpStart UI or carrying out programmatically through the SageMaker Python SDK. Let's explore both techniques to help you choose the method that finest suits your requirements.
Deploy DeepSeek-R1 through SageMaker JumpStart UI
Complete the following steps to deploy DeepSeek-R1 using SageMaker JumpStart:
1. On the SageMaker console, select Studio in the navigation pane.
2. First-time users will be triggered to produce a domain.
3. On the SageMaker Studio console, select JumpStart in the navigation pane.
The model internet browser displays available designs, with details like the company name and model abilities.
4. Look for DeepSeek-R1 to see the DeepSeek-R1 design card.
Each design card reveals crucial details, consisting of:
- Model name
- Provider name
- Task classification (for example, Text Generation).
Bedrock Ready badge (if relevant), showing that this model can be registered with Amazon Bedrock, allowing you to utilize Amazon Bedrock APIs to conjure up the model
5. Choose the design card to see the design details page.
The design details page includes the following details:
- The model name and setiathome.berkeley.edu company details. Deploy button to deploy the design. About and Notebooks tabs with detailed details
The About tab includes essential details, such as:
- Model description. - License details.
- Technical specifications.
- Usage standards
Before you deploy the design, it's suggested to examine the design details and license terms to validate compatibility with your use case.
6. Choose Deploy to proceed with deployment.
7. For Endpoint name, use the automatically generated name or develop a custom one.
- For example type ¸ pick an (default: ml.p5e.48 xlarge).
- For Initial instance count, go into the variety of circumstances (default: 1). Selecting appropriate instance types and counts is essential for cost and performance optimization. Monitor your release to change these settings as needed.Under Inference type, Real-time reasoning is picked by default. This is enhanced for sustained traffic and low latency.
- Review all setups for accuracy. For this model, we strongly recommend adhering to SageMaker JumpStart default settings and making certain that network isolation remains in place.
- Choose Deploy to release the model.
The implementation procedure can take several minutes to finish.
When deployment is complete, your endpoint status will alter to InService. At this point, the design is ready to accept inference demands through the endpoint. You can monitor the release progress on the SageMaker console Endpoints page, which will display appropriate metrics and status details. When the implementation is complete, you can invoke the design utilizing a SageMaker runtime customer and incorporate it with your applications.
Deploy DeepSeek-R1 utilizing the SageMaker Python SDK
To get going with DeepSeek-R1 using the SageMaker Python SDK, you will require to install the SageMaker Python SDK and make certain you have the required AWS permissions and environment setup. The following is a detailed code example that shows how to deploy and utilize DeepSeek-R1 for inference programmatically. The code for deploying the design is supplied in the Github here. You can clone the notebook and range from SageMaker Studio.
You can run extra requests against the predictor:
Implement guardrails and run reasoning with your SageMaker JumpStart predictor
Similar to Amazon Bedrock, you can also use the ApplyGuardrail API with your SageMaker JumpStart predictor. You can create a guardrail utilizing the Amazon Bedrock console or the API, and execute it as displayed in the following code:
Clean up
To prevent unwanted charges, complete the actions in this area to clean up your resources.
Delete the Amazon Bedrock Marketplace deployment
If you deployed the design utilizing Amazon Bedrock Marketplace, complete the following actions:
1. On the Amazon Bedrock console, under Foundation designs in the navigation pane, choose Marketplace implementations. - In the Managed releases area, locate the endpoint you desire to delete.
- Select the endpoint, and on the Actions menu, select Delete.
- Verify the endpoint details to make certain you're erasing the appropriate release: 1. Endpoint name.
- Model name.
- Endpoint status
Delete the SageMaker JumpStart predictor
The SageMaker JumpStart design you released will sustain expenses if you leave it running. Use the following code to erase the endpoint if you want to stop sustaining charges. For more details, see Delete Endpoints and Resources.
Conclusion
In this post, we explored how you can access and deploy the DeepSeek-R1 design using Bedrock Marketplace and SageMaker JumpStart. Visit SageMaker JumpStart in SageMaker Studio or Amazon Bedrock Marketplace now to get going. For more details, refer to Use Amazon Bedrock tooling with Amazon SageMaker JumpStart models, SageMaker JumpStart pretrained models, Amazon SageMaker JumpStart Foundation Models, Amazon Bedrock Marketplace, and Beginning with Amazon SageMaker JumpStart.
About the Authors
Vivek Gangasani is a Lead Specialist Solutions Architect for Inference at AWS. He assists emerging generative AI business develop innovative options utilizing AWS services and accelerated calculate. Currently, he is concentrated on developing techniques for fine-tuning and optimizing the inference performance of large language models. In his downtime, Vivek delights in hiking, enjoying films, and attempting various foods.
Niithiyn Vijeaswaran is a Generative AI Specialist Solutions Architect with the Third-Party Model Science group at AWS. His area of focus is AWS AI accelerators (AWS Neuron). He holds a Bachelor's degree in Computer technology and Bioinformatics.
Jonathan Evans is an Expert Solutions Architect dealing with generative AI with the Third-Party Model Science team at AWS.
Banu Nagasundaram leads product, engineering, and tactical collaborations for Amazon SageMaker JumpStart, SageMaker's artificial intelligence and generative AI hub. She is enthusiastic about developing services that help consumers accelerate their AI journey and unlock business value.