「training」の検索結果

重みが公開されたLLMの回答拒否に関わる仕組みを分析し、再学習なしで重みを変更する実験を支援。計算環境に合うモデルや手法の選択、変更前後の品質評価まで扱います。

• LLMの回答拒否の仕組みを調べたいとき
• 再学習せず重み変更を実験したいとき
• 計算環境に合うモデルと手法を選びたいとき

長い入力を扱うTransformerの学習・推論で、注意機構の計算を高速化し、GPUメモリ使用量を抑えるスキル。導入から性能測定、出力の比較まで案内します。

• 長い入力での学習を高速化したいとき
• 長文推論のメモリ使用量を抑えたいとき
• 処理速度とモデル出力の比較

限られたGPUメモリで大規模言語モデルを追加学習するため、LoRAやQLoRAの設定を支援します。学習結果の保存、用途別の切り替え、モデルへの統合も扱います。

• 少ないGPUメモリで追加学習したいとき
• LoRAの学習設定を選びたいとき
• 用途別のアダプターを切り替えたいとき

Train RuView models — camera-free WiFlow pose (10 sensor signals, no labels), camera-supervised pose (MediaPipe + ESP32 CSI → 92.9% PCK@20, ADR-079), RuVector contrastive embeddings (AETHER, ADR-024), domain generalization (MERIDIAN, ADR-027), local SNN environment adaptation, plus GPU training on GCloud and Hugging Face publishing. Use when building, fine-tuning, evaluating, or shipping a model.

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。

Fetch up-to-date documentation and code examples for any library, framework, SDK, CLI tool, or cloud service. Use whenever the user asks about a specific library — even well-known ones like React, Next.js, Prisma, Express, Tailwind, Django, or Spring Boot — because training data may not reflect recent API changes or version updates. Always use for: API syntax questions, configuration options, version migration issues, "how do I" questions mentioning a library name, debugging that involves library-specific behavior, setup instructions, and CLI tool usage. Use even when you think you know the answer. Do not rely on training data for API details, signatures, or configuration options — they are frequently out of date. Prefer this over web search for library 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。

Retrieves up-to-date documentation, API references, and code examples for any developer technology. Use this skill whenever the user asks about a specific library, framework, SDK, CLI tool, or cloud service — even for well-known ones like React, Next.js, Prisma, Express, Tailwind, Django, or Spring Boot. Your training data may not reflect recent API changes or version updates. Always use for: API syntax questions, configuration options, version migration issues, "how do I" questions mentioning a library name, debugging that involves library-specific behavior, setup instructions, and CLI tool usage. Use even when you think you know the answer — do not rely on training data for API details, signatures, or configuration options as they are frequently outdated. Always verify against current docs. Prefer this over web search for library documentation and API details.

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。

Supports work with Outpost Bio's open microbiome foundation models - the Waypoint checkpoints (Waypoint-6m, Waypoint-45m, Waypoint-170m), the Atlas pretraining corpus, the Compass eight-task benchmark, or the `waypoint` CLI from the `waypoint-bio` package. Covers embedding microbiome samples, fine-tuning on taxonomic abundance data, benchmarking a checkpoint on Compass, pretraining a GPT-2 model on taxonomic abundance profiles, and converting MetaPhlAn, Kraken2, QIIME 2, or MGnify abundance tables into waypoint format.

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。

Deep learning framework (PyTorch Lightning / lightning package). Organize PyTorch code into LightningModules, configure Trainers for multi-GPU/TPU, implement data pipelines, callbacks, logging (W&B, TensorBoard, MLflow), distributed training (DDP, FSDP, DeepSpeed), for scalable neural network training.

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。

Convert evaluation traces and production logs into SFT examples and preference pairs. Use when graded traces or failure examples exist and need to become training data, when applying rejection sampling to model outputs, or when building DPO pairs from passing and failing runs.

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。

Build end-to-end MLOps pipelines from data preparation through model training, validation, and production deployment. Use when creating ML pipelines, implementing MLOps practices, or automating model training and deployment workflows.

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。

Manage unified memory and thermals during long-running ML jobs on NVIDIA DGX Spark. Use when planning memory headroom for a training run on GB10, when a job OOMs on unified memory, or when monitoring temperature and power during multi-hour training.

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。

Identify poisoned training data and backdoored ML models across the pipeline using IBM's Adversarial Robustness Toolbox (activation clustering, spectral signatures, trigger reconstruction), Cleanlab for label-quality issues, and supply-chain checks like weight-hash verification and safetensors enforcement. Use before training or deploying on third-party/user-contributed data or downloaded checkpoints, during ML supply-chain reviews, or when investigating model misbehavior tied to specific inputs (suspected backdoor trigger).

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。

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。

Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Use when you want clean RL abstractions, easy algorithm experimentation, or scalable training with Monarch and TorchTitan.

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。

Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. 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。

Visualize training metrics, debug models with histograms, compare experiments, visualize model graphs, and profile performance with TensorBoard - Google's ML visualization toolkit

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。

Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable features, analyzing superposition, or studying monosemantic representations in language models.

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。

Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization

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。

Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond single models, or experimenting rapidly with model variants. Covers SLERP, TIES-Merging, DARE, Task Arithmetic, linear merging, and production deployment 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。

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.

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。

/cs:cdo-review <plan> — Decision-driven Chief Data Officer interrogation of any plan that touches training data, data architecture, data productization, or data team hiring. Use when validating training-data rights before model work, choosing warehouse vs lakehouse vs mesh, or valuing data assets for productization or M&A.

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。

Chief Data Officer advisory for startups: AI training data rights and consent provenance, data product strategy (warehouse vs lakehouse vs mesh, build-vs-buy), B2B customer-data-as-asset valuation and M&A readiness, data team org evolution. Use when deciding whether to train models on customer data, choosing data architecture, valuing data for fundraising or M&A, sequencing data hires, or when user mentions CDO, chief data officer, data strategy, data mesh, lakehouse, training data, data product, data monetization, or customer data asset. NOT a tactical data engineering skill — strategic decisions only.

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。

Diagnose GKE Cloud TPU multi-slice training hangs (Megascale `HANG_DETECTED` logs and hang monitored events) using the ML Diagnostics `Megascale XLA (MXLA) Hang Analyzer` (`gcloud alpha mldiagnostics monitored-events`) and 1-minute Cloud Monitoring multi-slice latency metrics (`kubernetes.io/container/multislice/*`). Distinguishes XLA compiler/HLO launch divergence and host data-input stalls from TPU chip, SparseCore, ICI, or network fabric faults. Use when multi-slice TPU training jobs freeze without progressing steps, emit `HANG_DETECTED`, or stall in collective operations. Don't use for gradual step-time throughput drops without hangs (use gke-ai-troubleshooting-tpu-performance-degradation) or pod preemption/eviction restarts (use gke-ai-troubleshooting-jobset-interruption).

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。

Interactively monitor training metrics from the current Codex session, periodically checking WandB or fallback logs for NaN, divergence, plateaus, and broken runs.

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。