Skill UI

This skill manages the entire lifecycle of long-running deep learning training and inference jobs deployed on remote, ephemeral computing environments (e.g., AutoDL, RunPod, bare SSH, Slurm, K8s). It provides critical capabilities like resilient checkpointing, managing spot instance preemption, debugging resource failures (OOM, NaN), and ensuring safe billing teardown, making it essential for professional ML Ops on rented resources.

This skill provides comprehensive methods and frameworks to detect sophisticated AI attacks, including model extraction, membership inference, and model inversion, which abuse inference APIs. It focuses on monitoring per-principal query volume, input distribution, and confidence exposure. Use it for securing public or partner inference APIs, or for conducting pre-deployment red-teaming to assess the model's extractability and overall security posture against MITRE ATLAS techniques.

A comprehensive guide and skill for deploying large language models (LLMs) locally using the llama.cpp framework and GGUF quantization format. It covers searching the Hugging Face Hub, selecting optimal quantizations (e.g., Q4_K_M), running local servers, and achieving OpenAI-compatible API endpoints. Supports various hardware backends including CPU, Mac Metal, CUDA, and ROCm. Ideal for developers needing offline, private LLM inference.

Automates the creation and publication of polished, functional Gradio demos on Hugging Face Spaces for any user-provided LoRA model. This skill handles the complex workflow of designing the optimal UI, selecting the correct inference pipeline (e.g., text-to-image, video-to-video), and generating all necessary files (app.py, requirements.txt, README.md) to deploy a complete, ready-to-use AI playground.

This guide helps authors position their machine learning systems manuscripts for the rigorous Conference on Machine Learning and Systems (MLSys). It details the venue's specific focus on system-level contributions, including training/inference infrastructure, compilers, data systems, and hardware-aware ML. Learn how to frame your work to appeal to systems reviewers by emphasizing measurable impacts on capability, cost, reliability, or deployment scale.

A comprehensive guide for academic authors detailing how to establish and maintain rigorous reproducibility in Machine Learning research papers (e.g., ICLR). It instructs on mapping every central claim to verifiable evidence, including recording seeds, variance, specific compute budgets (training vs. inference), data preprocessing scripts, and ethical considerations. The statement must function as a public contract for the scientific community.

This skill processes unstructured, rambling, stream-of-consciousness input (like dictation or brainstorming) into a highly structured 'echo' audit. It captures the user's core mission, locked decisions, open questions, and all model inferences, ensuring the user verifies the model's belief, not just the stated doubts. Ideal for refining fuzzy ideas before committing to work.

A specialized guide for authors aiming to submit to the Conference on Uncertainty in Artificial Intelligence (UAI). It provides expert advice on framing the core contribution—such as probabilistic reasoning, causal inference, and uncertainty modeling—and calibrating the manuscript for highly technical, statistical ML reviewers. Use this skill to refine your submission strategy, identify evidence gaps, and properly position your work relative to other top-tier AI venues like AISTATS and IJCAI.

Utilize zero-shot NER to extract custom entity types (e.g., Drug, Symptom, Device) from complex clinical or biomedical text without requiring any labeled training data or fine-tuning. OpenMed wraps GLiNER/GLiNER2, allowing users to define labels on the fly. The workflow involves building a model index and then running inference via CLI or Python API, making it ideal for rapid prototyping or unstable schemas.

This document outlines the canonical contract for deploying model inference services on Itô compute. It details the secure handoff process of an immutable serving manifest after a compute booking is verified. The process enforces multiple strict gates, including entitlement checks, manifest integrity validation, idempotency key usage, and final health checks, ensuring robust and reliable deployment of advanced AI models.

Orchestrate AI/ML workflows including data ingestion, model training, batch inference, and RAG indexing using tools like Prefect, Airflow, or Dagster. Essential for managing complex AI systems and automating ML pipelines.

Build Java applications with Azure AI Anomaly Detector SDK for univariate and multivariate time-series anomaly detection, streaming detection, change point detection, model training, and batch inference.