запрос активен

Поиск skills

Найдите подходящую возможность для вашего агента.

ground-truth-management

Comprehensive guide to creating, managing, and maintaining ground truth datasets for AI evaluation including annotation, quality control, and versioning

AmnadTaowsoam

data-ai

open

machine-learning

train

Execute a neural network training run with mandatory monitoring and best-practice defaults. Use when user wants to train a model, start training, or run a training job.

rHedBull

data-ai

open

machine-learning

scaling-analysis

Run scaling experiments to understand model/data/compute relationships. Use when investigating scaling laws, compute-optimal training, or model size decisions.

rHedBull

data-ai

open

machine-learning

Follow these patterns when implementing quant domain resources like Dataset, Signal, Alpha, Portfolio, Strategy, Universe, Backtest, or MonitoringRun in OptAIC. Use for creating DB models, DTOs, services, and tests for trading-specific entities.

colingwuyu

data-ai

open

framework-internals

performance-optimization

Optimize training performance for Kaggle competitions. Use PROACTIVELY when training is slow, GPU memory issues occur, or user asks about performance tuning. Keywords: 遅い, 最適化, OOM, CUDA, batch_size, num_workers, 高速化, パフォーマンス

kigasudayooo

development

open

machine-learning

feature-extraction

Extracts vector embeddings for fairness analysis using MobileNetV3 with FairDisCo disentanglement

do-ops885

data-ai

open

machine-learning

mlops-patterns

Follow these patterns when implementing MLOps features in OptAIC. Use for ML model definitions (5-component structure), model instances, training/inference pipelines, model registry, and monitoring. Covers signal models, macro regime models, relevance models, and signal combining/filtering models.

colingwuyu

data-ai

open

machine-learning

network-architecture-sizing

PPO network architecture sizing for trading models. Trigger: (1) model files are unexpectedly small/large, (2) choosing hidden_dims for training, (3) balancing model capacity vs inference speed.

smith6jt-cop

data-ai

open

machine-learning

refactorscikit-learn

Refactor Scikit-learn and machine learning code to improve maintainability, reproducibility, and adherence to best practices. This skill transforms working ML code into production-ready pipelines that prevent data leakage and ensure reproducible results. It addresses preprocessing outside pipelines, missing random_state parameters, improper cross-validation, and custom transformers not following sklearn API conventions. Implements proper Pipeline and ColumnTransformer patterns, systematic hyperparameter tuning, and appropriate evaluation metrics.

SnakeO

data-ai

open

machine-learning

training-archive-gating

Mandatory training archive with model gating (APPROVED/REVIEW/DROP). Trigger when: (1) training run completes, (2) need to decide which models to deploy, (3) want historical training reference, (4) need checkpoint recommendations for overfitting.

smith6jt-cop

data-ai

open

machine-learning

dpo

Direct Preference Optimization for learning from preference pairs. Covers DPOTrainer, preference dataset preparation, implicit reward modeling, and beta tuning for stable preference learning without explicit reward models. Includes thinking quality patterns.

atrawog

data-ai

open

machine-learning

training-improvements-v245

Training improvements: LR warmup, validation intervals, reward weights. Trigger when: (1) training unstable in early epochs, (2) need more validation visibility, (3) model too conservative.

smith6jt-cop

data-ai

open

machine-learning

pytorch-lightning

High-level training framework for PyTorch that abstracts boilerplate while maintaining flexibility. Includes the Trainer, LightningModule, and support for multi-GPU scaling and reproducibility. (lightning, pytorch-lightning, lightningmodule, trainer, callback, ddp, fast_dev_run, seed_everything)

cuba6112

data-ai

open

machine-learning

marker-engine-rl

Vertieft den Marker-Engine-Skill um SFT/RL-Feinabstimmung mit LeanDeep 4.0; lädt Marker aus Supabase/ZIP und lernt eine Policy zur präzisen, kontextualisierten Marker-Anwendung bei strikter Bottom-up-Logik.

DYAI2025

data-ai

open

machine-learning

ai-engineering-skill

Practical guide for building production ML systems based on Chip Huyen's AI Engineering book. Use when users ask about model evaluation, deployment strategies, monitoring, data pipelines, feature engineering, cost optimization, or MLOps. Covers metrics, A/B testing, serving patterns, drift detection, and production best practices.

odewahn

data-ai

open

machine-learning

training-resilience

Fix PPO training early-stop issues. Trigger when: (1) impossible drawdown values (>100%), (2) training stops too early, (3) need adaptive recovery instead of hard stop.

smith6jt-cop

data-ai

open

machine-learning

trl

This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.