Claude Skills by VectorSpaceLab

"Extend MMDeploy with rewriters, backend/codebase support, custom

"Use MMDeploy SDK model directories, runtime FFIs, profiler output,

"Evaluates exported MMDeploy models, profiles latency, runs

"Use this skill when working with MMDetection 3.x for object

"Choose, inspect, modify, and validate MMDetection 3.3.0 configs

"Extend MMDetection 3.3.0 with custom registries, models, losses,

"Prepare and validate MMDetection datasets, annotation formats,

"Use MMDetection 3.3.0 public APIs for image, batch, folder,

"Plan and validate MMDetection 3.3.0 training, resume, testing,

"Use MMDetection3D for 3D detection, segmentation, dataset

Select, inspect, and adapt MMDetection3D configs and model-zoo entries.

"Customize MMDetection3D datasets, data pipelines, model

"Prepare MMDetection3D datasets, custom dataset layouts, conversion

"Run and adapt MMDetection3D point-cloud, monocular,

Package, publish, inspect, and troubleshoot MMDetection3D serving utilities.

"Work with MMDetection3D box, point, coordinate-conversion, and

"Construct and troubleshoot MMDetection3D training, testing,

"Use MMEngine to configure, train, evaluate, inspect, visualize,

"Create, validate, merge, dump, and debug MMEngine

"Design, validate, and troubleshoot MMEngine datasets, transforms,

"Implement and troubleshoot MMEngine model, metric, evaluator,

"Build, adapt, and troubleshoot MMEngine Runner and FlexibleRunner

"Use MMEngine logging, message hubs, visualizer backends,

"Routes MMGeneration tasks across sampling, training, evaluation,

"Guides MMGeneration latent editing, interpolation, projection,

"Guides MMGeneration config authoring, dataset layout, registry

"Guides MMGeneration evaluation commands, metric selection, and

"Routes MMGeneration pretrained-model loading, sampling,

"Guides MMGeneration training, resume, distributed launch, and

"Use MMOCR for OCR inference, dataset preparation, config-driven

"Prepare, inspect, and troubleshoot MMOCR dataset conversion and

"Use and extend MMOCR registries, model components, data samples,

"Use MMOCR inference APIs and safe CLI preflights for text

"Route MMOCR config-driven training, testing, evaluation, model-zoo

"Operate MMPreTrain model-zoo inference, config-driven training and

"Route dataset layouts, annotation validation, pipeline design, and

"Operate MMPreTrain model zoo discovery, inference APIs, task

"Route post-training analysis, visualization, checkpoint

"Inspect MMPreTrain configs, plan train/test commands, and choose

"Use MMSegmentation for semantic segmentation configs, datasets,

"Configure MMSegmentation configs, datasets, transforms, class

"Run MMSegmentation semantic segmentation and depth inference from

"Customize MMSegmentation models, registries, projects,

"Launch, troubleshoot, and inspect MMSegmentation training,

"Guides MMSkeleton skeleton-based action-recognition, skeleton-data

"It prepares, validates, loads, normalizes, transforms, inspects,

"Provides optional image and video pose-estimation workflows with

"This skill provides ST-GCN skeleton action-recognition model,

"Use MMYOLO for OpenMMLab YOLO object detection configs, datasets,

"Choose, customize, and validate MMYOLO model-family configs