9,677 件 ・ 関連度順
概要と使いどころ
Comprima modelos de linguagem grandes usando destilação de conhecimento de modelos professor para aluno. Use ao implantar modelos menores com desempenho retido, transferir capacidades do GPT-4 para modelos de código aberto ou reduzir custos de inferência. Aborda escalamento de temperatura, alvos suaves, KLD reversa, destilação de logits e estratégias de treinamento MiniLLM.
日本語の概要は準備中です。原文の説明を表示しています。
Detect model stealing, model inversion, and membership inference performed through inference-API abuse by monitoring query patterns, applying output perturbation, and red-teaming your own model's extractability.
日本語の概要は準備中です。原文の説明を表示しています。
Use CodexBar CLI local cost usage to summarize per-model usage for Codex or Claude, including the current (most recent) model or a full model breakdown. Trigger when asked for model-level usage/cost data from codexbar, or when you need a scriptable per-model summary from codexbar cost JSON.
日本語の概要は準備中です。原文の説明を表示しています。
Role-based multi-model strategy. The model catalog lives in a local configuration file instead of the skill text -- roles (thinker/decision-maker, long-running worker, draft generator, reviewer/acceptance, escalation), reasoning effort as an independent axis, mandatory review for draft models, provider switching for independent reviews, budgets, delegation flow, and cost/quality of small workers.
日本語の概要は準備中です。原文の説明を表示しています。
Orquestación multimodelo y estrategia de cambio de modelos. Selección de modelos basada en puntuación, delegación entre agentes (Gemini, Codex, Ollama), emparejamiento con advisor, disparadores de escalado, matriz de permisos y optimización de eficiencia de costos.
日本語の概要は準備中です。原文の説明を表示しています。
Rollenbasierte Multi-Modell-Strategie. Modellkatalog liegt in einer lokalen Konfigurationsdatei statt im Skill-Text -- Rollen (Denker/Entscheider, Dauer-Worker, Entwurfsgenerator, Reviewer/Abnahme, Eskalation), Reasoning-Effort als eigene Achse, Abnahmepflicht für Entwurfsmodelle, Anbieterwechsel für unabhängige Reviews, Kontingente, Delegationsablauf und Kosten/Qualität kleiner Worker.
日本語の概要は準備中です。原文の説明を表示しています。
Use when the user wants to send the same prompt to several LLMs and compare the answers - sequentially (stopwatch, clean per-lane timing) or in parallel (true race), with optional repetitions per model, judged by the starting model across quality, correctness, completeness, instruction fidelity and latency (time is only one dimension). Triggers on /compare-race, 'model race', 'vergleiche modelle', 'gleicher prompt an mehrere modelle'.
日本語の概要は準備中です。原文の説明を表示しています。
Repository-grounded threat modeling that enumerates trust boundaries, assets, attacker capabilities, abuse paths, and mitigations, and writes a concise Markdown threat model. Trigger only when the user explicitly asks to threat model a codebase or path, enumerate threats/abuse paths, or perform AppSec threat modeling. Do not trigger for general architecture summaries, code review, or non-security design work.
日本語の概要は準備中です。原文の説明を表示しています。
Pick the right serving container for a SageMaker model deployment and find its current image URI. Use this skill whenever about to deploy a model to a SageMaker endpoint and an image URI needs to be chosen — including when the user says "deploy this LLM", "host this HuggingFace model", "serve this fine-tuned model", "deploy this embedding model", "host a reranker", "serve a sentence-transformers model", or when about to hardcode any container URI in deployment code. HuggingFace-curated Deep Learning Containers are ALWAYS preferred: HuggingFace vLLM (LLMs and generative rerankers), HuggingFace vLLM-Omni (multimodal), TEI (embeddings/cross-encoder rerankers), HF Inference Toolkit (other transformers). Generic images (AWS vLLM, DJL-LMI, SGLang) are used only when no HuggingFace image is compatible — never merely because they carry a newer version. Never hardcode a container URI from memory and never default to TGI. Prevents stale-image failures and wrong-region URIs.
日本語の概要は準備中です。原文の説明を表示しています。
Integrate on-device AI using Foundation Models framework, Core ML, and open-source LLM runtimes on Apple Silicon. Covers Foundation Models (LanguageModelSession, @Generable, @Guide, SystemLanguageModel, structured output, tool calling), Core ML (coremltools, model conversion, quantization, palettization, pruning, Neural Engine, MLTensor), MLX Swift (transformer inference, unified memory), and llama.cpp (GGUF, cross-platform LLM). Use when building tool-calling AI features, working with guided generation schemas, converting models, or running on-device inference.
日本語の概要は準備中です。原文の説明を表示しています。
Integrate and optimize Core ML models in iOS apps for on-device machine learning inference. Covers model loading (.mlmodelc, .mlpackage), predictions with auto-generated classes and MLFeatureProvider, compute unit configuration (CPU, GPU, Neural Engine), MLTensor, VNCoreMLRequest, MLComputePlan, multi-model pipelines, and deployment strategies. Use when loading Core ML models, making predictions, configuring compute units, or profiling model performance.
日本語の概要は準備中です。原文の説明を表示しています。
Deploy a new LLM model via LiteLLM (GitOps) and/or Azure (Terraform + node backend), add it to the ai toolkit, create the required pull requests, and verify end-to-end. Use when rolling out a model to any environment, adding a model to the toolkit, troubleshooting "model not available" errors, or finding pricing/token limits/model ids for any provider.
日本語の概要は準備中です。原文の説明を表示しています。
ScalarDBレビューエージェント - 設計レビュー（--mode=design）とコードレビュー（--mode=code）を実行。/review-scalardb --mode=design で呼び出し。
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.
日本語の概要は準備中です。原文の説明を表示しています。
Hermes Agent CLI リファレンス。 インストール、CLI コマンド (chat, model, gateway, config, skills, cron, webhook, mcp, profile)、 Configuration (Terminal Backends, Memory, TTS / STT, Compression, Display, Auxiliary Models)、 Features (Tools, Skills, MCP, Memory, Voice Mode, Personality, Context Files)、 Messaging Gateway (14+ プラットフォーム)、Security、Architecture、 Platform Support、Nous Portal 連携 (Tool Gateway)。
Apple データ永続化・データフロー・ネットワーク・ユーティリティリファレンス。 SwiftData / Core Data / Foundation / Combine。 @Model、ModelContainer、ModelContext、@Query、Predicate、FetchDescriptor、 NSManagedObject、NSFetchRequest、NSPersistentContainer、NSFetchedResultsController、 URLSession、JSONEncoder、JSONDecoder、Publisher、@Published、AnyCancellable、sink。
Facilitate Kimball-style dimensional (data warehouse / BI) modeling sessions through conversation, producing DimML (Dimensional Modeling Language, YAML) as the primary artifact plus a bus-matrix + star-schema HTML. Use whenever the user wants to design fact tables, dimensions, a star schema, or a bus matrix — declaring grain, choosing SCD types, identifying conformed dimensions and measures (additivity) — by grilling one step at a time. Also invoke for refining an existing DimML, or when the user says "ディメンショナル・モデリング", "ディメンションモデリング", "スタースキーマ", "バスマトリクス", "ファクト表を設計したい", "グレインを決めたい", "DimML を育てたい", "Kimball で整理したい".
Analyzes time-series and longitudinal RNA-seq for differential expression and trajectory structure. Covers DESeq2 LRT with reduced models, time as factor vs continuous vs natural splines, maSigPro (Nueda 2014 for RNA-seq), ImpulseDE2 with explicit impulse-model failure modes, DREAM for repeated measures via linear mixed models, pseudoreplication avoidance, conditional vs marginal modeling, and trajectory clustering with DPGP, Mfuzz (with Schwämmle 2010 fuzzifier estimation), and splines+k-means. Use when modeling time-course or longitudinal expression, choosing factor vs spline, handling repeated measures from the same subject, avoiding pseudoreplication, clustering temporal trajectories, or selecting between dedicated time-course tools and pairwise+LRT.
日本語の概要は準備中です。原文の説明を表示しています。
Calls ChIP-seq peaks with MACS3, MACS2, HOMER, or SPP across narrow (TF) and broad (histone) modes. Handles input control matching, fragment-size modeling vs --nomodel, effective genome size, ENCODE-style IDR vs naive overlap, hyper-ChIPable artifacts, and aligner-specific shifts. Use when calling peaks from ChIP-seq alignments, choosing between narrow vs broad mode for a histone mark, deciding model vs nomodel for low-depth data, applying ENCODE pseudoreplicate IDR, or reconciling MACS vs HOMER vs SPP results.
日本語の概要は準備中です。原文の説明を表示しています。
Analyzes time-series and longitudinal RNA-seq for differential expression and trajectory structure. Covers DESeq2 LRT with reduced models, time as factor vs continuous vs natural splines, maSigPro (Nueda 2014 for RNA-seq), ImpulseDE2 with explicit impulse-model failure modes, DREAM for repeated measures via linear mixed models, pseudoreplication avoidance, conditional vs marginal modeling, and trajectory clustering with DPGP, Mfuzz (with Schwämmle 2010 fuzzifier estimation), and splines+k-means. Use when modeling time-course or longitudinal expression, choosing factor vs spline, handling repeated measures from the same subject, avoiding pseudoreplication, clustering temporal trajectories, or selecting between dedicated time-course tools and pairwise+LRT.
日本語の概要は準備中です。原文の説明を表示しています。
Calls ChIP-seq peaks with MACS3, MACS2, HOMER, or SPP across narrow (TF) and broad (histone) modes. Handles input control matching, fragment-size modeling vs --nomodel, effective genome size, ENCODE-style IDR vs naive overlap, hyper-ChIPable artifacts, and aligner-specific shifts. Use when calling peaks from ChIP-seq alignments, choosing between narrow vs broad mode for a histone mark, deciding model vs nomodel for low-depth data, applying ENCODE pseudoreplicate IDR, or reconciling MACS vs HOMER vs SPP results.
日本語の概要は準備中です。原文の説明を表示しています。
LLM APIの費用を抑えるため、作業の複雑さに応じたモデル選択、予算の記録、一時的なエラーの再試行、共通プロンプトのキャッシュを組み合わせる設計例を示します。
AIによる開発の成功条件を実装前に定義し、コード・ルール・モデル・人の評価で新機能と既存機能を確認しながら、複数回の試行から信頼性を記録するスキル。
言語モデルに指示への応答や好ましい回答を学習させるため、TRLによる追加学習を案内するスキル。データ準備、学習方式の選択、報酬モデルの学習や評価を扱います。