README.md

metadata

language:
  - zh
  - en
tags:
  - dia-pangu
  - pangu
  - large-language-model
  - dialogue
  - chinese
  - english
library_name: transformers
pipeline_tag: text-generation
model_type: pangu
license: other

Dia-Pangu

Dia-Pangu is a domain-adapted large language model trained based on the OpenPangu-7B architecture.
This repository provides the trained model weights for both Chinese and English versions.

📦 Model Weights

This repository contains two sets of model weights:

• CN/pytorch_model.bin — Chinese version
• EN/pytorch_model.bin — English version

These weights were trained as part of our Dia-Pangu project.

For full training details, architecture, and usage instructions, please refer to the official GitHub repository:

👉 https://github.com/chunyu-atx/Dia-Pangu

---

🚀 Usage

The detailed inference and fine-tuning instructions are provided in the GitHub repository:

https://github.com/chunyu-atx/Dia-Pangu

Please follow the setup instructions there to properly load and run the model.

---

📖 Model Description

The core functions of Dia-Pangu are CT image recognition and generation of Chinese and English reports.\ This model was trained using the CTRG-Chest-548K dataset and underwent Lora fine-tuning.

⚠️ Important Notes

• This repository only contains model weights.
• The full model implementation and loading scripts are available in the GitHub repository.
• Make sure to use the correct tokenizer and configuration files provided in the main project repository.