opensite-ai/opensite-skills / aws-ai-ml
aws-ai-ml
VerifiedSelects, deploys, and customizes AI models on Amazon SageMaker. Training or Processing jobs, fine-tuning (SFT/DPO/RLVR/RLAIF), model selection, dataset preparation, evaluation, SageMaker or Bedrock deployment, inference optimization and endpoint diagnostics. Covers the full lifecycle from planning through production. Use when fine-tuning models on SageMaker, choosing which base model to customize, fine-tune, or deploy from SageMaker JumpStart or Hub, SageMakerPublicHub, or the SageMaker public model catalog, transforming or validating training data, evaluating model quality, deploying or optimizing endpoints, configuring IAM/S3 for training, or managing SageMaker Managed MLflow. Use for endpoint health, failures, latency, logs, metrics, errors. Covers Serverless Model Customization, Nova and OSS deployment paths, and PySDK v3. NOT for Ground Truth labeling, Feature Store, or general-purpose AWS infrastructure.
references/model-deployment/references/optimization/recommendation-options.md
View on GitHub# Recommendation Configuration Options
## Instance Types (optional)
For **latency** and **throughput** targets, the user can specify which instance types to evaluate:
> "Would you like to specify which instance types to evaluate, or let the service choose automatically?
>
> Examples: `ml.g5.xlarge`, `ml.g6.12xlarge`, `ml.p5.48xlarge`"
Note: **Cost** target does not support customer-specified instance types.
## Optimization (optional)
> "Should the service try to optimize the model?
>
> - **Kernel tuning** — Up to 30% improved TTFT and throughput
> - **Speculative decoding** — Lower latency via a draft model that predicts tokens ahead (requires a compatible draft model)
>
> **Heads up:** Without optimizations: 30–60 min. With: 2–10 hours.
>
> Default is yes (`advanced_optimization=True`). Set `advanced_optimization=False` to skip."
### Dataset Requirement for Throughput + Optimization
If **throughput** target AND `advanced_optimization=True`, a dataset is **required**. Without it, the job fails with a validation error.
Ask the user for a dataset (an S3 location), then **inspect it yourself** to figure out the format — don't quiz them on schema names:
> "Since you chose throughput with optimizations, I need a dataset to drive it. Point me at an S3 location — either a JSONL dataset of prompts, or your endpoint's SageMaker Data Capture prefix to replay real traffic. What's the S3 URI?"
⏸ Wait for user.
When the user provides the S3 URI, read a couple of records and identify the schema, then build the workload in Cell 3 of `../../code_templates/optimize-recommendation.py` with `Workload.from_dataset(...)` instead of `Workload.synthetic(...)`.
`custom_dataset_type` is an **optional hint** naming the dataset's schema (not required — omit to let AIPerf auto-detect). See `benchmarking-guidance.md` → "Dataset schema hint (`custom_dataset_type`)" for the schema → value mapping and the dataset-transformation fallback for unsupported formats.
The `Workload.from_dataset(...)` alternative is **Cell 3 of `../../code_templates/optimize-recommendation.py`** (a commented-out block in that cell) — swap it in for `Workload.synthetic(...)` and fill `[DATASET_S3_URI]` and `custom_dataset_type` (from the schema you detected, or omit for auto-detection). Generate it from the template rather than writing it inline.
## Workload Config
Required. Defined as Cell 3 of the notebook (from `../../code_templates/optimize-recommendation.py`). `generate_deployment_recommendations` takes the `Workload` object via `workload=` and **auto-creates** the underlying `AIWorkloadConfig` from it. `workload_config_name=` is the **name to give that newly-created config** (not a reference to a pre-existing one) — the template passes the agent-generated `WORKLOAD_CONFIG_NAME` from Cell 2; if you omit it, the SDK auto-generates a name. Do not pass a config name into `workload=` — that parameter expects a `Workload` object.
## Inference Framework (optional)
> "Which inference framework?
>
> - **VLLM** — High-performance serving with PagedAttention
> - **LMI** — SageMaker Large Model Inference container
>
> Default: auto-selected based on the model."
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |