Scenario: A team needs to fine-tune an LLM for text summarization using a low- code/no-code (LCNC) solution to automate model training and minimize manual intervention. Question- Which solution will best meet the team’s requirements?. Options:

Amazon SageMaker JumpStart is the ideal solution for the team's requirements. It provides a low-code/no-code (LCNC) environment within SageMaker Studio to fine-tune pre-trained foundation models, including those for text summarization. Users can select a model, provide their training dataset through a simple UI, configure hyperparameters, and launch the fine-tuning job with minimal manual intervention. This process automates the underlying infrastructure setup and training orchestration, directly aligning with the need for an automated, LCNC solution. While SageMaker Autopilot is typically for tabular data, the spirit of the option points to the automated, streamlined fine-tuning experience that SageMaker JumpStart provides for LLMs.

B. Fine-tuning on raw Amazon EC2 instances, even via SageMaker Studio, is a code-intensive process requiring manual environment setup, data management, and script development, which contradicts the LCNC requirement.

C. SageMaker Training Jobs are a powerful, code-first solution. They require users to write and manage training scripts and configurations, which is the opposite of a low-code/no-code approach.

D. SageMaker Script Mode is a feature for running custom training scripts. This approach offers high flexibility but demands significant coding effort, violating the core LCNC and minimal intervention requirements.

1. Amazon SageMaker Developer Guide: Under the section "Amazon SageMaker JumpStart," the documentation states, "With SageMaker JumpStart, you can also fine-tune models using a no-code user interface in SageMaker Studio or programmatically through the SageMaker Python SDK." This confirms the availability of a no-code path for fine-tuning.

Source: AWS Documentation, Amazon SageMaker Developer Guide, "Amazon SageMaker JumpStart".

2. Amazon SageMaker Developer Guide: The section "Fine-tune a foundation model" provides a step-by-step walkthrough of the UI-based process. It details how to "Choose a pre-trained model to fine-tune," "Configure the training job," and "Launch the training job" all within the SageMaker Studio graphical interface, highlighting the minimal coding required.

Source: AWS Documentation, Amazon SageMaker Developer Guide, "Get started with foundation models," -> "Fine-tune a foundation model".

3. Amazon SageMaker Developer Guide: The documentation for SageMaker Training Jobs and Script Mode ("Train a Model with Amazon SageMaker") clearly outlines a process that involves providing your own training algorithm code or script, which is fundamentally different from an LCNC approach.

Source: AWS Documentation, Amazon SageMaker Developer Guide, "Train a Model with Amazon SageMaker".