ConardLi/easy-dataset

Project overview

An application that converts domain-specific documents in formats such as PDF, Markdown, DOCX, TXT, and EPUB into structured datasets for LLM fine-tuning, RAG, and model evaluation. It provides multiple text splitting algorithms and a human blind test system for model performance and RAG recall rate assessment.

Project type

RAG · Data Processing · Evaluation & Observability

Use cases

Data Analysis

Deployment

Refer to project documentation

License

License pending

Best for

• AI engineers and data teams preparing structured datasets from text-based domain documents for fine-tuning, retrieval, or multi-dimensional evaluation.

Key capabilities

• Parses PDF, Markdown, DOCX, TXT, EPUB and additional formats with intelligent recognition.
• Provides multiple splitting algorithms including Markdown structure, recursive separators, fixed length, and code-aware chunking, with customizable visual segmentation.
• Creates structured datasets from documents applicable for fine-tuning, RAG, and evaluation.
• Converts domain-specific documents into evaluation test sets and automatically runs multi-dimensional evaluation tasks.
• Meets needs such as vertical domain model evaluation, post-fine-tuning model performance assessment, and RAG recall rate evaluation.

Limitations and risks

• The application specifies text modalities; support for audio, video, or image inputs is not indicated.
• Operating system support, minimum hardware requirements, necessary databases, external services, and GPU requirements are not documented.

Getting started

• Setup difficulty, first success paths, and initial configuration instructions are not present in the available facts.

Evidence and sources

• GitHub project description: A powerful tool for creating datasets for LLM fine-tuning 、RAG and Eval
• README: Easy Dataset is an application specifically designed for building large language model (LLM) datasets. It features an intuitive interface, along with built-in powerful document pa…
• README: Easy Dataset Version 1.7.0 launches brand-new evaluation capabilities! You can effortlessly convert domain-specific documents into evaluation datasets (test sets) and automaticall…
• README: Intelligent Document Processing: Supports PDF, Markdown, DOCX, TXT, EPUB and more formats with intelligent recognition
• Release: [1.7.3] 2026-04-09

Reviewed at: 2026-07-14T23:29:36Z