Construction-PPE fine-tuning with audited data splits, held-out evaluation, and PyTorch/independent ONNX deployment
Safe, auditable YOLO missing-label recovery | 安全可审计的 YOLO 漏标恢复工具
Quality-aware, read-only MCP server for discovering and selecting African speech corpora, Wolof first, with audited metrics, provenance, licensing, and training-set planning.
Audit clinical NLP datasets before training. Offline, deterministic, no cloud model.
Reproducible n-gram screening of language-model training data for overlap with evaluation benchmarks.
Petroleum-engineer-friendly coverage audit of the Equinor Volve dataset from a frozen filesystem catalog.
Reproducible audit of the MAD dataset (MAST failure taxonomy): three undocumented taxonomy versions, renumbered codes, and what that means for reported inter-annotator agreement.
Local-first QA and leakage/drift auditor for medical image segmentation datasets.
Companion repository for annotations, back translations, benchmark analyses, and supporting materials from our audit of RWTH-PHOENIX-2014T as a benchmark for German Sign Language translation.
只读审计你的 LoRA 训练数据集：caption 遗漏 / 重复 / 触发词漂移 / 扩展名与实际格式不符。零依赖，出一个能双击打开的 HTML。
Deterministic entity-by-source coverage audits for multi-source datasets.
NeurIPS 2026 E&D Accept (Poster) · Matched-budget audit of recaptioned image-text supervision for text-to-image training
To associate your repository with the dataset-audit topic, visit your repo's landing page and select "manage topics."