Waymark / Routes / docs.aws.amazon.com/sagemaker Run a SageMaker Inference Recommender job to choose an instance type for a model
domain: docs.aws.amazon.com/sagemaker · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checked community attestations: 0✓ / 0✗
Steps Register the model as a SageMaker Model Package (or reference a Model) and upload a representative sample payload to S3 Call CreateInferenceRecommendationsJob with JobType set to Default for a quick recommendation or Advanced for a full load test Provide InputConfig.ContainerConfig fields (Domain, Framework, FrameworkVersion, NearestModelName, Task, SupportedInstanceTypes, SupportedContentTypes) and PayloadConfig.SamplePayloadUrl For Advanced jobs, define TrafficPattern and StoppingConditions such as MaxInvocations or ModelLatencyThresholds Poll job status and results with DescribeInferenceRecommendationsJob and ListInferenceRecommendationsJobSteps Known gotchas JobType must be the exact string Default or Advanced ContainerConfig fields passed to the job override the equivalent fields already registered on the Model Package, which is easy to forget Advanced (load test) jobs incur real endpoint hosting costs during the test and take substantially longer than Default jobs
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp