> For the complete documentation index, see [llms.txt](https://low-codeai.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://low-codeai.gitbook.io/docs/how-it-works/decentralized-model-training/distributed-data-contribution.md).

# Distributed Data Contribution

In the decentralized training model, users can contribute data to the training process from various sources, whether it's from their personal datasets, business data, or third-party data providers. This distributed approach allows the model to be trained on a wider variety of data, improving its generalization and making it more robust to different use cases. Participants contribute data without revealing sensitive information, ensuring privacy and security through blockchain.

* **Data Privacy:** Since the data remains decentralized, users retain control over their own datasets, with encryption and blockchain ensuring that sensitive information is not exposed.
* **Data Variety:** A more diverse range of data inputs enhances the model’s ability to generalize, improving its performance across different scenarios.