More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Frameworks, Managed Models and Model Selection

    1

    Master PyTorch, Hugging Face, cloud managed models and selection criteria. 10 ordered skills, starting with Edge AI Inference, ONNX Conversion and Interoperability, AutoML.

    Free

    Security scanned
    Frameworks, Managed Models and Model Selection

    Frameworks, Managed Models and Model Selection

    Example session with this skill installed

    A small team at an invented company, "Northwind Sensors", wants a PyTorch classifier that labels vibration windows from machines as normal or faulty. Their loss turns NaN after epoch 3 when they enable float16 mixed precision, and two runs with the same code give different validation scores.

    • Read your context and instructions
    • Compiled the frameworks managed models

    PyTorch Training Report: machine fault classification (fictional)

    Setup: group split by machine (train 84 / val 18 / test 18 machines), 1D CNN, lr 1e-3, batch 128, 20 epochs, 1 GPU

    Sanity checks: 8-batch overfit passed; initial loss matched expectation after class weighting

    Reproducibility settings: seeds for torch/random/NumPy, DataLoader generator and worker_init_fn, deterministic algorithms on; caveat: not guaranteed identical across versions or devices

    Results: val F1 curves stable after fixes (no NaN across 3 seeds); test F1 reported once; logistic baseline 0.61 compared; improvement stated with the 3-seed spread

    Issues found and fixes: window-level leakage; gradients cleared per epoch; missing eval mode and no_grad; hand-written log loss overflow under float16

    Checkpoint contents and resume: model, optimizer, scheduler, GradScaler state, epoch, config; resume tested by one extra epoch

    Next steps: hand the model to hyperparameter-tuning; export check with onnx-conversion-and-interoperability

    Connects securely to your tools. The creator never sees your data.

    About this skill

    The problem

    Engineers and consultants choosing the AI stack often need a repeatable process, not a one-off prompt. This pack targets: Master PyTorch, Hugging Face, cloud managed models and selection criteria.

    What's inside (10 skills, suggested learning order)

    1. Edge AI Inference — Run models on local devices.
    2. ONNX Conversion and Interoperability — Export models across frameworks.
    3. AutoML — Automate model selection and tuning.
    4. Hyperparameter Tuning — Systematically optimize hyperparameters.
    5. Deep Learning with PyTorch — Build neural networks in PyTorch.
    6. Hugging Face Transformers — Use and adapt pretrained models.
    7. PySpark for Large-Scale ML — Train and process massive data.
    8. Databricks and Lakehouse for ML — Unified data and ML platform.
    9. Managed Cloud Models (Bedrock, Azure OpenAI, Vertex) — Consume ready-made models through managed platforms.
    10. Model Selection and Evaluation by Use Case — Choose the right model by cost, quality and latency.

    Why a skill pack instead of a one-off prompt

    Each skill gives the agent a defined trigger, step-by-step procedure and output format, so results are consistent across sessions and easier to review.

    How to use

    1. Install the pack in your SKILL.md-compatible agent.
    2. Ask for the task in plain language (for example: "run models on local devices").
    3. Review the output against the skill's quality checks before using it.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.

    Outputs are drafts that need human review; they are not professional, legal or financial advice. Tools, APIs and prices change; check current vendor documentation before relying on version-specific details. Quality depends on the context you give the agent (data, requirements, constraints). Instruction-only skills: they produce documents and plans, not running software.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 8 days ago

    • Free with an Agensi account

    Permissions

    No special permissions declared or detected

    Listed8 days ago

    What's inside

    Frequently Asked Questions

    10 skills, each with SKILL.md and references/sources.md: Edge AI Inference; ONNX Conversion and Interoperability; AutoML; Hyperparameter Tuning; Deep Learning with PyTorch; Hugging Face Transformers; PySpark for Large-Scale ML; Databricks and Lakehouse for ML; Managed Cloud Models (Bedrock, Azure OpenAI, Vertex); Model Selection and Evaluation by Use Case.

    Engineers and consultants choosing the AI stack.

    Agents that read the SKILL.md format. See the Compatibility note for what was actually tested.

    New in LLM Ops

    Free