datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ConvoDrift_human_eval_conversational_dataset
ConvoDrift Human Evaluation Data
This folder contains human evaluations of ConvoDrift conversations from three
annotators. Each conversation has six prompt-response pairs, quality ratings
for up to Q1-Q8, and optional corrections to drift and direction labels.
Files
Configuration
Records
Description
annotator_1
500
Complete evaluation file for Annotator 1
annotator_2
500
Complete evaluation file for Annotator 2
annotator_3
500
Complete evaluation… See the full description on the dataset page: https://huggingface.co/datasets/Vihindi-K/ConvoDrift_human_eval_conversational_dataset.myanmar_quran_parallel_dataset_human_vs_ai
Myanmar Quran Parallel Dataset: Human vs AI
This dataset is a comprehensive multi-parallel corpus of the Holy Qur'an, containing all 6,236 verses.
It is designed as a high-quality linguistic resource for evaluating and aligning AI systems on formal, literary, and modern Myanmar (Burmese) language in a religious context.
Each verse aligns the original Uthmani Arabic text with trusted human translations and multiple AI-generated translations, enabling fine-grained comparison between… See the full description on the dataset page: https://huggingface.co/datasets/freococo/myanmar_quran_parallel_dataset_human_vs_ai.human_curated_qa_dataset
Human Curated QA Dataset
DigiGreen/human_curated_qa_dataset is a human-verified question-answer dataset designed to support research and development in natural language question answering and agriculture-focused conversational AI.
This dataset contains realistic, domain-relevant QA pairs that were manually curated to ensure accurate and contextually rich answers. It can be used to benchmark models for QA generation.
📌 Dataset Overview
Name: Human Curated QA Dataset… See the full description on the dataset page: https://huggingface.co/datasets/DigiGreen/human_curated_qa_dataset.