dataset-config
Routes NanoDet dataset, preprocessing, and config-validation workflows for COCO, XML, YOLO, and model-assembly setups.
- License 许可
- Apache-2.0
- 最近更新
- 2026-10-08
- Stars
- ★ 11
- Forks
- 3
01 前言 Front Matter
- name
- dataset-config
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Routes NanoDet dataset, preprocessing, and config-validation workflows for COCO, XML, YOLO, and model-assembly setups.
- disable-model-invocation
- true
02 资源文件 3
- 参考文档references/configuration.md3.3 KB
- 参考文档references/data-formats.md2.2 KB
- 参考文档references/troubleshooting.md1.1 KB
03 SKILL.md 正文
dataset-config
Use this sub-skill when you need to understand or validate a NanoDet config, prepare a dataset layout, or check the preprocessing path before training or inference.
Use this route for
- Reading or editing YAML config files.
- Verifying
save_dir,model,data,device,schedule,evaluator,log, andclass_namesfields. - Choosing between
CocoDataset,XMLDataset, andYoloDataset. - Checking
keep_ratio,multi_scale, and augmentation pipeline settings. - Confirming that a config can build the expected model family.
Do not use this route for
- Training / validation / checkpoint lifecycle logic. Use
traininginstead. - Image/video/webcam inference or model export. Use
inference-exportinstead. - Repository maintenance or packaging chores.
Read first
references/configuration.mdfor the NanoDet config structure and field meanings.references/data-formats.mdfor COCO, XML, and YOLO data-layout details.references/troubleshooting.mdfor config, dataset, and preprocessing failures.../../references/api-reference.mdwhen you need verified builder signatures.
Skill-owned scripts
scripts/check_config.py— load a config, check class counts, and build the model as a smoke test.scripts/check_dataset.py— build a dataset from a config and inspect a sample.
Typical workflow
- Load the YAML config.
- Check the model/dataset fields that affect runtime behavior.
- Verify the dataset layout and annotation format.
- Run the skill-owned smoke script before moving to training or inference.
Cross-links
- If the config is valid but training fails later, switch to
training. - If the config is valid and you only need to run a demo or export a model, switch to
inference-export.
04 同仓库更多 Skills
obsidian-clipper-template-creator
Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.
obsidian-clipper-template-creator
Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.
python-api
Programmatic Object-Detection-Metrics API guidance for BoundingBox, BoundingBoxes, Evaluator, enums, coordinate conversion, AP/mAP semantics, and source-style import diagnostics.
file-evaluation
Folder-based PASCAL VOC AP/mAP evaluation from ground-truth and detection text files.
object-detection-metrics
Guides legacy Object-Detection-Metrics PASCAL VOC AP/mAP evaluation from detection text folders and direct Python API classes.
nx-workspace-patterns
Configure and optimize Nx monorepo workspaces. Use when setting up Nx, configuring project boundaries, optimizing build caching, or implementing affected commands.