825 件 ・ 関連度順
概要と使いどころ
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.
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。
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.
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。
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.
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。
Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT. Use when compressing models without retraining, achieving 50% sparsity with minimal accuracy loss, or enabling faster inference on hardware accelerators. Covers unstructured pruning, structured pruning, N:M sparsity, magnitude pruning, and one-shot methods.
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。
Selects biomarker features from high-dimensional omics data using Boruta all-relevant selection, mRMR, LASSO/elastic-net, and stability selection, while controlling the leakage, irreproducibility, and correlated-feature traps that make most published signatures fail to replicate. Use when identifying candidate biomarkers, deciding between an all-relevant and a minimal-optimal selector, or judging whether a selected gene set is reproducible. For unbiased performance estimation of the resulting model see machine-learning/model-validation; for interpreting a trained model see machine-learning/prediction-explanation.
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。
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.
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。
Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limited, need to fit larger models, or want faster inference. Supports INT8, NF4, FP4 formats, QLoRA training, and 8-bit optimizers. Works with HuggingFace Transformers.
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。
Implement minimal network events and component state with authority, conversion, dirtying, and regression coverage.
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。
Structured bug diagnosis and fixing workflow — reproduce, diagnose root cause, apply minimal fix, write regression test, and scan for similar patterns
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。
MUST USE for deploying Aerospike on Kubernetes. Contains CE-specific YAML templates, validated AerospikeCluster CR examples, and critical constraints that prevent enterprise-only config mistakes (feature-key-file, security sections crash CE pods). Without this skill, deployments fail on first attempt due to CE 8.1 breaking changes (data-size not memory-size, no info port 3003) or webhook map/list shape rules (service/network must be maps; logging must be a list). Triggers on: deploy/create/set up Aerospike on K8s, kind, minikube, EKS, GKE; AerospikeCluster CR; ACKO operator; spec.operations / WarmRestart / PodRestart YAML; NoSQL database on Kubernetes. 9 ready-to-use YAML examples from minimal single-node to full-featured multi-rack.
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。
Reflection-first problem solving methodology for Convex development. Use before implementing any solution to ensure proper root cause analysis, 98% code confidence, and minimal change scope.
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。
This skill should be used when the user asks to "create AGENTS.md", "update AGENTS.md", "maintain agent docs", "set up CLAUDE.md", or needs to keep agent instructions concise. Enforces research-backed best practices for minimal, high-signal agent documentation.
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 a coding task should be driven end-to-end from issue intake through implementation, review, deployment, and acceptance verification with minimal human re-intervention.
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。
Graceful, refined aesthetic with delicate typography, minimal palettes, and polished layouts that exude sophistication.
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。
Two-dimensional minimalist style with vibrant colors, clean typography, and no 3D effects for fast, user-friendly interfaces.
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。
Shadcn/ui-inspired design with minimal, clean components, monochrome palette, and utility-first patterns.
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。
Modern minimalist aesthetic with clean lines, intentional color palette, subtle interactions, and consistent spacing.
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。
Stripped-back design emphasizing whitespace, clean typography, and restrained color for maximum clarity and focus.
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。
Paper-textured, print-inspired design with minimal colors, clean serif/sans typography, and tactile surface qualities.
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。
Carefully curated, modern minimal style with elegant serif typography and understated, sophisticated palettes.
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。
Contemporary editorial style with serif typography, minimal palettes, and clean layouts for polished digital products.
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。
Playful, minimal design with bright colors, rounded shapes, tactile 3D borders, and friendly illustrations for approachable interfaces.
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。
A cyber-slick, dark-only Matrix-inspired interface defined by minimalist fashion, high-tech digital elements
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。
Raw, anti-design aesthetic inspired by concrete architecture with unadorned elements, jarring layouts, and functional minimalism.
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。