「train」の検索結果

Track and visualize ML training experiments with Trackio. Use when logging metrics during training (Python API), firing alerts for training diagnostics, or retrieving/analyzing logged metrics (CLI). Supports real-time dashboard visualization, alerts with webhooks, HF Space syncing, and JSON output for automation.

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。

Understand the constraints a design was built under and evaluate quality within that context. Use when reviewing work that had limited time, budget, or team size, or when assessing whether constraints were handled well.

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。

Set up and run multi-turn RL training for interactive environments (terminal tasks, tool use, search/RAG, games) using the Tinker API. Use when the user wants multi-turn RL, agentic training, tool-use RL, or interactive environment 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。

Implement GCP Organization Policy constraints to enforce security guardrails across the entire resource hierarchy, restricting risky configurations and ensuring compliance at organization, folder, and project levels.

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。

Z3 SMT solver ile CTF constraint çözme — byte equations, hash collision, validator bypass

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。

Cost modeling and ROI analysis for specialized LLM development. Use when deciding whether to train a custom model, estimating total cost, or calculating break-even vs frontier APIs. Covers training costs, inference costs, and time-to-ROI projections.

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 Fully Sharded Data Parallel training with PyTorch FSDP - parameter sharding, mixed precision, CPU offloading, FSDP2

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。

Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model 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。

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

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 distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/FP8, 1-bit Adam, sparse attention

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。

Training patterns for autoregressive neural PDE solvers (FNO, DeepONet, CNO). Covers rollout training, noise injection for stability, multi-component loss functions (H1, frequency-sensitive, boundary-aware), per-channel normalization for coupled multi-variable systems, and the PDEBench nRMSE metric. Use when training any neural operator that predicts time-dependent PDE solutions.

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。

GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII redaction, NSFW detection. Scales across GPUs with RAPIDS. Use for preparing high-quality training datasets, cleaning web data, or deduplicating large corpora.

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 DPO training pipelines for AI alignment. TRIGGERS - Use when user needs help with ai-dpo-trainer related tasks.

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。

Design accessibility training. TRIGGERS - Use when user needs help with accessibility-training-program related tasks.

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 you need to design, review, or improve validation in Quarkus applications — including Bean Validation on JAX-RS resources, @Valid on parameters and CDI beans, constraint groups, @ConfigMapping validation, custom constraints, nested DTO validation, and ExceptionMapper-based error mapping. This should trigger for requests such as Add validation support in Quarkus; Review Quarkus validation rules; Improve request validation in Quarkus REST APIs; Add custom validation constraints in Quarkus; Validate Quarkus @ConfigMapping properties. Part of cursor-rules-java project

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。

Assigns pathogen lineages (SARS-CoV-2 Pangolin via UShER mode; Nextclade clade + QC; pango-designation alias_key.json resolution) and tracks variant frequencies over time using Nextstrain (Augur + Auspice), wastewater deconvolution (Freyja, COJAC, alcov, lineagespot), lineage fitness modelling (Wenseleers / Bedford-Figgins multinomial logistic), and recombinant detection (3SEQ, RDP4, Bolotie). Covers Pangolin pangolin-data version pinning (mandatory for reproducibility), Nextclade dataset versioning (lineage-defining mutations change with dataset), Freyja barcode forward-only date constraint, ARTIC primer scheme version churn (V3 / V4 / V4.1 / V5.3.2 / Midnight 1200) with documented dropout regions, recombinant X-prefix Pango designation lag, GISAID vs INSDC dual-deposition tensions, and the Karthikeyan 2022 wastewater early-detection signal with explicit reproducibility caveats. Use when assigning Pango lineages and Nextclade clades to viral consensus sequences, building Nextstrain Augur surveillance pipelines, deconvolving wastewater pooled samples into lineage frequencies with Freyja, tracking lineage frequencies and growth advantages over time, pinning pangolin-data / Nextclade dataset versions for reproducibility, handling ARTIC primer dropouts (V4.1 amplicons 64 / 76 / 88-90), or running variant surveillance for SARS-CoV-2 / influenza / Mpox / RSV / H5N1 / measles.

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。

Assigns pathogen lineages (SARS-CoV-2 Pangolin via UShER mode; Nextclade clade + QC; pango-designation alias_key.json resolution) and tracks variant frequencies over time using Nextstrain (Augur + Auspice), wastewater deconvolution (Freyja, COJAC, alcov, lineagespot), lineage fitness modelling (Wenseleers / Bedford-Figgins multinomial logistic), and recombinant detection (3SEQ, RDP4, Bolotie). Covers Pangolin pangolin-data version pinning (mandatory for reproducibility), Nextclade dataset versioning (lineage-defining mutations change with dataset), Freyja barcode forward-only date constraint, ARTIC primer scheme version churn (V3 / V4 / V4.1 / V5.3.2 / Midnight 1200) with documented dropout regions, recombinant X-prefix Pango designation lag, GISAID vs INSDC dual-deposition tensions, and the Karthikeyan 2022 wastewater early-detection signal with explicit reproducibility caveats. Use when assigning Pango lineages and Nextclade clades to viral consensus sequences, building Nextstrain Augur surveillance pipelines, deconvolving wastewater pooled samples into lineage frequencies with Freyja, tracking lineage frequencies and growth advantages over time, pinning pangolin-data / Nextclade dataset versions for reproducibility, handling ARTIC primer dropouts (V4.1 amplicons 64 / 76 / 88-90), or running variant surveillance for SARS-CoV-2 / influenza / Mpox / RSV / H5N1 / measles.

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 this skill when documenting supervisor reasonable-suspicion training for Part 382 compliance. Covers what to log, what to retain, and what training providers to use.

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 this skill to evaluate which defensive driving program is best for your fleet. Compares Smith System, CarriersEdge, Compass-native training, and other options.

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 this skill when selecting a TPR-registered Training Provider for behind-the-wheel ELDT training. Covers the registry, evaluation criteria, and common pitfalls.

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 carrier needs to build an internal dispatcher training program, onboard new dispatchers, or certify dispatcher competency. Covers job description, required skills, training modules, certification frameworks (CDS, CTP), shadowing, evaluation, and career path.

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 this skill when implementing or renewing supervisor reasonable-suspicion training. Covers the 60-minute alcohol + 60-minute drug training requirement, recordkeeping, and refresher cadence.

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 PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.

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。