Get started fine-tuning foundation models in Amazon SageMaker Unified Studio
Amazon SageMaker Unified Studio provides a large collection of state-of-the-art foundation models. These models support use cases such as content writing, code generation, question answering, copywriting, summarization, classification, information retrieval, and more. You can find and deploy these foundation models in the JumpStart model catalog. In some cases, you can also customize them. You can use the foundation models to build your own generative AI solutions for a wide range of applications.
A foundation model is a large pre-trained model that is adaptable to many downstream tasks and often serves as the starting point for developing more specialized models. Examples of foundation models include Meta Llama 4 Maverick 17B, DeepSeek-R1, or Stable Diffusion 3.5 Large. These models are pre-trained on massive amounts of data.
Model customization
You might need to customize a base foundation model to better align it with your specific use cases. The recommended way to first customize a foundation model is through prompt engineering. Providing your foundation model with well-engineered, context-rich prompts can help achieve desired results without any fine-tuning or changing of model weights. For more information, see Prompt engineering for foundation models in the Amazon SageMaker AI Developer Guide.
If prompt engineering alone is not enough to customize your foundation model to a specific task, you can fine-tune a foundation model on additional domain-specific data. The fine-tuning process involves changing model weights.
To help you learn how to fine-tune foundation models, Amazon SageMaker Unified Studio provides an example training dataset for each model that's eligible for training. You can also choose to fine-tune the model with your own data set. Before you can do that, you must prepare your data set and store it in an Amazon S3 bucket. The required format for the data set varies between models. You can learn about the required format in the model details page in Amazon SageMaker Unified Studio.
Fine-tuning a foundation model
One way to fine-tune a model in Amazon SageMaker Unified Studio is to use JumpStart. First, you browse the model catalog to find a model that's eligible for fine-tuning. Then, you train the model with a training data set. Follow these steps to learn how to fine-tune with this approach.
-
Sign in to Amazon SageMaker Unified Studio using the link that your administrator gave you.
-
Choose a model to train by doing the following:
-
From the main menu, choose Build.
-
From the drop-down menu, under Model Development, choose Jumpstart Models.
-
If the Select or create project to continue window appears, select a project that you've created, and choose Continue.
The JumpStart page lists the model providers.
-
Choose a provider to see the available models.
Not all providers have models that you can fine-tune in JumpStart. If you want to quickly find an eligible model so that you can get familiar with fine-tuning, choose Meta. It has many trainable models to choose from.
Tip
For some providers, you can filter the list of models so that you see only the trainable ones. Choose the Trainable checkbox if it's present.
-
From the provider's list of models, choose the model you want to train.
Amazon SageMaker Unified Studio shows the model details page, which provides information from the model provider. If you want to prepare a custom fine-tuning data set, use this page to learn the required format.
-
-
From the model details page, if the model is trainable, choose Train to create a training job.
If the model isn't trainable, the button is disabled. In that case, return to the JumpStart page, find a different model that's trainable, and try again.
-
On the Fine-tune model page, under Artifacts, do one of the following:
-
Keep the default selection of Example training dataset. This dataset is useful when you want to learn how to fine-tune with Amazon SageMaker Unified Studio. However, it won't be effective for customizing the model for your specific needs.
-
If you've prepared a custom training dataset, choose Enter training dataset, and provide the URI that locates it in Amazon S3.
-
-
For Output artifact location (S3 URI), specify where Amazon SageMaker Unified Studio uploads the fine-tuned model. You can choose to use the default bucket, or you can specify a custom location in Amazon S3.
-
(Optional) Under Hyperparameters, update the hyperparameters you want to change.
The hyperparameters available for each trainable model differ depending on the model. Review the help text and additional information in the model details pages in Amazon SageMaker Unified Studio to learn more about hyperparameters specific to the model of your choice.
For more information on available hyperparameters, see Commonly supported fine-tuning hyperparameters in the Amazon SageMaker AI Developer Guide.
-
Under Compute, for Training Instance, specify the training instance type for your training job. You can choose only from instances that are compatible with the chosen model.
Important
Choose an instance type that fits within the service quotas for your AWS account.
When you submit your training job, Amazon SageMaker Unified Studio attempts to provision the chosen instance type in your account. This attempt succeeds only if your quotas have remaining capacity for the instance type.
To see the quotas for your account, open the Service Quotas console at https://console.aws.amazon.com/servicequotas/
. If you want to use a specific instance type but lack the required quota capacity, you can request a quota increase with Support. For more information, see Requesting a quota increase in the Service Quotas User Guide.
-
(Optional) Under Information, for Training Job Name, you can edit the default name.
-
(Optional) For Tags, you can add and remove tags in the form of key-value pairs to help organize and categorize your fine-tuning training jobs.
-
Enter Submit to submit the training job.
Note
Some models require acceptance of an end-user license agreement (EULA). If this applies to the model that you choose to fine-tune, Amazon SageMaker Unified Studio prompts you with a window that contains the EULA content. You are responsible for reviewing and complying with any applicable license terms and making sure they are acceptable for your use case before using the model.
Amazon SageMaker Unified Studio shows a page with details about the training job. Here, you can observe the status of the job as it executes.
The training job might take a long time to complete. You can view it at any time from the Training jobs page.
When the training job completes, the status becomes Completed. After the job completes, you can choose Deploy to deploy the fine-tuned model to an inference endpoint.
