diff --git a/README.en.md b/README.en.md
new file mode 100644
index 0000000000000000000000000000000000000000..0207de13ac93a534815364784ce3fa97447f87cf
--- /dev/null
+++ b/README.en.md
@@ -0,0 +1,85 @@
+
+
+# ZhiQing Language Model Project
+
+ZhiQing is a Chinese language model based on the Transformer architecture, supporting multiple training modes including pre-training, instruction fine-tuning (SFT), reinforcement learning with human feedback (RLHF), and direct preference optimization (DPO). This project aims to develop a model capable of understanding and generating high-quality Chinese text.
+
+## Features
+
+- Supports multiple training modes: pre-training, SFT, RLHF, DPO  
+- Efficient distributed training and memory monitoring  
+- Custom Chinese tokenizer and model architecture  
+- Mixed precision training support  
+
+## Components
+
+- **Model Architecture**: Transformer-based decoder structure  
+- **Tokenizer**: Supports special processing for Chinese characters  
+- **Dataset**: Supports data formats for pre-training, SFT, DPO, and RLHF  
+- **Training Tools**: Includes learning rate scheduling, mixed precision training, distributed training, etc.  
+- **Evaluation Tools**: Includes GPU memory monitoring and distributed memory monitoring  
+
+## Installation
+
+1. Clone the repository:
+```bash
+git clone https://gitee.com/starboot/ai.git
+```
+
+2. Install dependencies:
+```bash
+pip install -r requirements.txt
+```
+
+## Usage
+
+### Initialize Model and Tokenizer
+
+```python
+from starboot.model_starboot import ZhiQingModel, ZhiQingConfig
+from starboot.token.starboot_tokenizer import ZhiQingTokenizer
+
+config = ZhiQingConfig()
+model = ZhiQingModel(config)
+tokenizer = ZhiQingTokenizer('path/to/vocab.txt')
+```
+
+### Train the Model
+
+```python
+from train import train
+
+train(num_epochs=10, batch_size=32)
+```
+
+### Inference
+
+```python
+input_text = "这是一个示例输入"
+input_ids = tokenizer.encode(input_text, return_tensors='pt')
+output = model.generate(input_ids, max_new_tokens=50)
+print(tokenizer.decode(output[0]))
+```
+
+### Supported Training Modes
+
+- Pre-training: `pretrain(num_epochs, batch_size)`  
+- Instruction Fine-Tuning: `sft_tuning(num_epochs, batch_size)`  
+- DPO Training: `dpo_tuning(num_epochs, batch_size)`  
+
+## Datasets
+
+The following dataset formats are supported:
+
+- Pre-training dataset: `ZhiQingPretrainDataset`  
+- Instruction Fine-Tuning dataset: `ZhiQingSftDataset`  
+- DPO dataset: `ZhiQingDpoDataset`  
+- RLHF dataset: `ZhiQingRlhfDataset`  
+
+## Contribution
+
+Contributions of code and suggestions are welcome. Please submit a Pull Request or open an Issue.
+
+## License
+
+This project is licensed under the MIT License. Please refer to the LICENSE file for details.
\ No newline at end of file
diff --git a/README.md b/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..99b28ec4b1110995487a9c0b6f7f5bd0307651d0
--- /dev/null
+++ b/README.md
@@ -0,0 +1,83 @@
+# ZhiQing 语言模型项目
+
+ZhiQing 是一个基于Transformer架构的中文语言模型，支持多种训练模式，包括预训练、指令微调(SFT)、基于人类反馈的强化学习(RLHF)和直接偏好优化(DPO)。该项目旨在开发一个能够理解和生成高质量中文文本的模型。
+
+## 特性
+
+- 支持多种训练模式：预训练、SFT、RLHF、DPO
+- 高效的分布式训练和内存监控
+- 自定义的中文分词器和模型架构
+- 混合精度训练支持
+
+## 组件
+
+- **模型架构**: 基于Transformer的解码器结构
+- **分词器**: 支持中文字符的特殊处理
+- **数据集**: 支持预训练、SFT、DPO和RLHF数据格式
+- **训练工具**: 包括学习率调度、混合精度训练、分布式训练等
+- **评估工具**: 包括GPU内存监控和分布式内存监控
+
+## 安装
+
+1. 克隆仓库:
+```bash
+git clone https://gitee.com/starboot/ai.git
+```
+
+2. 安装依赖:
+```bash
+pip install -r requirements.txt
+```
+
+## 使用
+
+### 初始化模型和分词器
+
+```python
+from starboot.model_starboot import ZhiQingModel, ZhiQingConfig
+from starboot.token.starboot_tokenizer import ZhiQingTokenizer
+
+config = ZhiQingConfig()
+model = ZhiQingModel(config)
+tokenizer = ZhiQingTokenizer('path/to/vocab.txt')
+```
+
+### 训练模型
+
+```python
+from train import train
+
+train(num_epochs=10, batch_size=32)
+```
+
+### 推理
+
+```python
+input_text = "这是一个示例输入"
+input_ids = tokenizer.encode(input_text, return_tensors='pt')
+output = model.generate(input_ids, max_new_tokens=50)
+print(tokenizer.decode(output[0]))
+```
+
+### 支持的训练模式
+
+- 预训练: `pretrain(num_epochs, batch_size)`
+- 指令微调: `sft_truning(num_epochs, batch_size)`
+- DPO训练: `dpo_truning(num_epochs, batch_size)`
+
+## 数据集
+
+支持以下数据集格式:
+
+- 预训练数据集: `ZhiQingPretrainDataset`
+- 指令微调数据集: `ZhiQingSftDataset`
+- DPO数据集: `ZhiQingDpoDataset`
+- RLHF数据集: `ZhiQingRlhfDataset`
+
+## 贡献
+
+欢迎贡献代码和建议。请提交Pull Request或Issue。
+
+## 许可证
+
+该项目使用MIT许可证。详情请查看LICENSE文件。
\ No newline at end of file