Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.

Write publication-ready ML/AI/Systems papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM, OSDI, NSDI, ASPLOS, SOSP. Use when drafting papers from research repos, structuring arguments, verifying citations, or preparing camera-ready submissions. Includes LaTeX templates, reviewer guidelines, and citation verification workflows.

Compress large language models using knowledge distillation from teacher to student models. Use when deploying smaller models with retained performance, transferring GPT-4 capabilities to open-source models, or reducing inference costs. Covers temperature scaling, soft targets, reverse KLD, logit distillation, and MiniLLM training strategies.

GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.

OpenAI's model connecting vision and language. Enables zero-shot image classification, image-text matching, and cross-modal retrieval. Trained on 400M image-text pairs. Use for image search, content moderation, or vision-language tasks without fine-tuning. Best for general-purpose image understanding.

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on consumer GPUs, when you need 4× memory reduction with <2% perplexity degradation, or for faster inference (3-4× speedup) vs FP16. Integrates with transformers and PEFT for QLoRA fine-tuning.

Activation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss. Use when deploying large models (7B-70B) on limited GPU memory, when you need faster inference than GPTQ with better accuracy preservation, or for instruction-tuned and multimodal models. MLSys 2024 Best Paper Award winner.

High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest neighbor search, hybrid search with filtering, or scalable vector storage with Rust-powered performance.

Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter tuning with Ray Tune, fault tolerance, elastic scaling. Use when training massive models across multiple machines or running distributed hyperparameter sweeps.

High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.

Review UI code for Web Interface Guidelines compliance. Use when asked to "review my UI", "check accessibility", "audit design", "review UX", or "check my site against best practices".

Fluid Framework client release group — minor releases, patch releases, and post-release type test updates. Covers release prep, branching, version bumps, changelogs, release notes, and type test baselines. In autonomous mode, auto-detects state from the schedule and repo, attempts to execute, and falls back to a GitHub issue on failure. Triggers on "release", "do the release", "release status", version bump, release notes, changelog, release branch, or release engineering.

This skill should be used when the user asks to "run policy check", "check policy", "policy-check", or needs to validate package compliance. Provides guidance on running policy checks for specific packages or the entire repository.

Trigger ADO pipelines for a Copilot-created PR by posting /azp run comments. Use when the user asks to trigger CI pipelines for a specific PR.

Generate comprehensive GitHub Feature Area Status reports for the Windows App SDK repository. Use when asked to create triage reports, identify high-priority issues, analyze feature area health, find issues needing attention, or generate status dashboards. Triggers on requests involving issue triage, area status, priority analysis, bug tracking reports, or engineering team focus areas.

Create and manage Git worktrees for parallel development workflows. Use when multiple self-contained issues should NOT be fixed in a single branch, when human-Copilot iteration requires isolated environments with separate chat history and commits, or when parallel work items need independent build/test results. Triggers on requests involving branch isolation, work item separation, parallel development, or avoiding messy branch switching.

Prepare weekly triage meeting summary for WinAppSDK Needs-Triage issues. Use when preparing for triage meetings, reviewing Needs-Triage issues, generating diff reports since last triage, summarizing new or updated issues, or creating action item recommendations. Triggers on requests involving triage preparation, Needs-Triage review, meeting summary, triage diff, or weekly issue analysis.

Query GitHub issues efficiently with jq argument support for filtering

Plan for adding temporary ID support to safe output jobs