## Getting Started
### Installation

**1. Prepare the code and the environment**

Git clone our repository, creating a python environment and ativate it via the following command

```bash
git clone https://github.com/DLYuanGod/ArtGPT-4.git
cd ArtGPT-4
conda env create -f environment.yml
conda activate artgpt4
```


**2. Prepare the pretrained Vicuna weights**

The current version of MiniGPT-4 is built on the v0 versoin of Vicuna-13B.
Please refer to our instruction [here](PrepareVicuna.md) 
to prepare the Vicuna weights.
The final weights would be in a single folder in a structure similar to the following:

```
vicuna_weights
├── config.json
├── generation_config.json
├── pytorch_model.bin.index.json
├── pytorch_model-00001-of-00003.bin
...   
```

Then, set the path to the vicuna weight in the model config file 
[here](minigpt4/configs/models/minigpt4.yaml#L16) at Line 16.

**3. Prepare the pretrained ArtGPT-4 checkpoint**
 [Downlad](https://huggingface.co/Tyrannosaurus/ArtGPT-4/blob/main/ArtGPT-4.pth)


Then, set the path to the pretrained checkpoint in the evaluation config file 
in [eval_configs/minigpt4_eval.yaml](eval_configs/minigpt4_eval.yaml#L10) at Line 11. 



### Launching Demo Locally

Try out our demo [demo.py](demo.py) on your local machine by running

```
python demo.py --cfg-path eval_configs/artgpt4_eval.yaml  --gpu-id 0
```


### Training
The training of ArtGPT-4 contains two alignment stages. The training process for the step is consistent with that of [MiniGPT-4](https://minigpt-4.github.io/).

**Datasets**
We use [Laion-aesthetic](https://github.com/LAION-AI/laion-datasets/blob/main/laion-aesthetic.md) from the LAION-5B dataset, which amounts to approximately 200GB for the first 302 tar files. 



## Acknowledgement

+ [MiniGPT-4](https://minigpt-4.github.io/) Our work is based on improvements to the model.



## License
This repository is under [BSD 3-Clause License](LICENSE.md).
Many codes are based on [Lavis](https://github.com/salesforce/LAVIS) with 
BSD 3-Clause License [here](LICENSE_Lavis.md).