Create README.md
Browse files# AutoMR for Pangu
This project equips the Pangu model with the AutoMR reasoning framework, optimized for Huawei Ascend hardware.
## 🌳 Project Structure
```
.
├── AMC.SH
├── automr
│ ├── config.py
│ ├── dag.py
│ ├── data_loader.py
│ ├── evaluator.py
│ ├── __init__.py
│ ├── model.py
│ ├── strategies.py
│ ├── trainer.py
│ └── utils.py
├── checkpoints
│ └── MATH
├── embedder_server.sh
├── generator_server.sh
├── main.py
├── math_train.sh
└── processed_data
├── AMC
└── MATH
```
## 🔧 1. Installation
### a. Clone the Project Repository
```bash
hf download Alexhf825/AutoMR-pangu --local-dir AutoMR-pangu
cd AutoMR-pangu
```
### b. Install Dependencies
```
```
### c. Download Datasets
This command will download the datasets and place them in the `./processed_data` directory, matching the project structure.
```bash
hf download Alexhf825/dataset-test-pangu --repo-type=dataset --local-dir=./processed_data
```
## 🚀 2. Start the Servers
This project requires two services running in an OpenAI-API style. Please run the following commands in **two separate terminal sessions**.
**Start the Embedder Server:**
```bash
bash embedder_server.sh
```
**Start the Generator Server:**
```bash
bash generator_server.sh
```
## 📈 3. Run Evaluation
A pre-trained checkpoint (`MATH`) is provided. You can directly evaluate the model on the `AMC` dataset using the following command:
```bash
bash AMC.sh
```
## 🏋️ 4. Run Training
You can also train the model from scratch on the `MATH` dataset by running:
```bash
bash math_train.sh
```