Zjh-819/LLMDataHub
A curated hub of high-quality datasets for LLM instruction finetuning and training.
Not currently ranked — collecting fresh signals.
star history
LLMDataHub aggregates open-source training corpora for large language models, covering alignment datasets, domain-specific datasets, pretraining corpora, and multimodal datasets. It provides links, size, language, usage guidance, and descriptions for each dataset to help researchers and developers train LLMs like Alpaca, Vicuna, and ChatGLM.
Frequently asked
- What is Zjh-819/LLMDataHub?
- A curated hub of high-quality datasets for LLM instruction finetuning and training.
- Is LLMDataHub open source?
- Yes — Zjh-819/LLMDataHub is open source, released under the MIT license.
- How popular is LLMDataHub?
- Zjh-819/LLMDataHub has 3.4k stars on GitHub.
- Where can I find LLMDataHub?
- Zjh-819/LLMDataHub is on GitHub at https://github.com/Zjh-819/LLMDataHub.