advanced-control
advanced-control
Route expert agents to STG/APG, Q8/VAE patching, latent normalization, decoder-noise control, attention and flow-edit tricks, PAG/FETA, RF samplers, inverse prediction, and utility nodes.
THE SKILLS DIRECTORY
スキルを探す.
AI ワークフローに合うスキルを見つけよう。
augmentation-and-sampling
augmentation-and-sampling
Guides braindecode signal transforms, composed augmentation loaders, and sequence, relative-positioning, and self-supervised samplers with shape-safe validation.
config-authoring
config-authoring
Design, inspect, validate, and serialize DataDesigner configs: builders, columns, samplers, seeds, person data, validators, processors, model/tool configs, custom columns, skip/drop rules, and Jinja references.
data-and-datasets
data-and-datasets
Operate FastReID dataset registries, built-in layouts, custom dataset classes, dataloaders, transforms, samplers, and safe data preflight checks.
data-designer
data-designer
Generate high-quality synthetic datasets using statistical samplers and Claude's native LLM capabilities. Use when users ask to create synthetic data, generate datasets, create fake/mock data, generate test data, training data, or any data generation task. Supports CSV, JSON, JSONL, Parquet output. Adapted from NVIDIA NeMo DataDesigner (Apache 2.0).
data-designer
data-designer
Generate high-quality synthetic datasets using statistical samplers and Claude's native LLM capabilities. Use when users ask to create synthetic data, generate datasets, create fake/mock data, generate test data, training data, or any data generation task. Supports CSV, JSON, JSONL, Parquet output. Adapted from NVIDIA NeMo DataDesigner (Apache 2.0).
data-designer
data-designer
Generate high-quality synthetic datasets using statistical samplers and Claude's native LLM capabilities. Use when users ask to create synthetic data, generate datasets, create fake/mock data, generate test data, training data, or any data generation task. Supports CSV, JSON, JSONL, Parquet output. Adapted from NVIDIA NeMo DataDesigner (Apache 2.0).
data-structures-and-io
data-structures-and-io
Design, validate, and troubleshoot MMEngine datasets, transforms, samplers, collate functions, data elements, and backend-agnostic file IO.
datasets-and-samplers
datasets-and-samplers
Use for TorchGeo GeoDataset/NonGeoDataset usage, dataset composition, raster/vector dataset implementation, sample dictionaries, collate functions, and patch samplers.
datasets-evaluation
datasets-evaluation
Prepare and validate MMDetection datasets, annotation formats, dataloaders, transforms, samplers, evaluators, and analysis workflows.
dpm-solver
dpm-solver
Use DPM-Solver and DPM-Solver++ single-file samplers for diffusion ODE sampling, PyTorch/JAX integration, ScoreSDE/DDPM examples, and Stable Diffusion acceleration.
evaluation-and-training
evaluation-and-training
Route sentence-transformers evaluation and training tasks across SentenceTransformer, CrossEncoder, and SparseEncoder trainers, losses, evaluators, dataset shapes, samplers, smoke tests, and safe training scaffolding.
generation-and-model-components
generation-and-model-components
Choose and configure Haystack prompt builders, generators, embedders, classifiers, samplers, validators, and provider/model-backed components.
imbalanced-learn
imbalanced-learn
Router for imbalanced-learn samplers, workflows, metrics, datasets, and balanced batch generators.
model-compatibility
model-compatibility
Model family compatibility matrix covering loaders, resolutions, samplers, CFG, VAE, ControlNet, and LoRA compatibility for SD 1.5, SDXL, Flux, SD3, and video models
models-and-customization
models-and-customization
Select RecBole model families and implement or diagnose custom RecBole models, trainers, dataloaders, samplers, and metrics.
optuna
optuna
Use this skill for Optuna hyperparameter optimization workflows: studies, trials, samplers, pruners, storage, CLI, visualization, artifacts, integrations, and advanced APIs.
patch-workflows
patch-workflows
Design TorchIO patch-based training queues and dense patch inference loops with samplers, SubjectsLoader batches, PatchLocation metadata, and PatchAggregator.
18 / 29
公開の技能源から直接収録
An independently imported catalog with original authors, licenses and source links. Classification is not a security audit.
Sources and import coverage