cysin/Vary

Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models

Haoran Wei*, Lingyu Kong*, Jinyue Chen, Liang Zhao, Zheng Ge, Jinrong Yang, Jianjian Sun, Chunrui Han, Xiangyu Zhang

Release

[2023/12/11] We released the online demo, have fun!
[2023/12/11] We released the codes of Vary (train and inference)!

Usage and License Notices: The data, code, and checkpoint are intended and licensed for research use only. They are also restricted to use that follow the license agreement of LLaMA, Vicuna, GPT-4, Qwen, and LLaVA.

Install
Vary Weights
Demo
Train

Install

Clone this repository and navigate to the Vary folder

git clone https://github.com/Ucas-HaoranWei/Vary.git
cd Vary

Install Package

conda create -n vary python=3.10 -y
conda activate vary
pip install e .

Install Flash-Attention

pip install ninja
pip install flash-attn --no-build-isolation

Vary Weights

Due to download speed issues with Baiduyun, we have temporarily closed the download link. Our weights will be reorganized and open source again in the next few days. If you are in urgent need of weights for your research recently, please contact me by email.
Download the CLIP-VIT-L in Hugging Face

Demo

Update the CLIP-VIT path in the codes (/cache/vit-large-patch14/) to your path.

python vary/demo/run_qwen_vary.py  --model-name  /vary/model/path/ --image-file /an/image/file.png

Train

We currently do not plan to open source the weights of the intermediate.
However, we release the train codes. So you can train on your own dataset. If you want to do this, you can try this:

For Vary-base (one machine, if you have multiple machines you need to prepare your host file)

deepspeed   Vary/train/train_qwen_vary.py  --deepspeed /Vary/zero_config/zero2.json
            --model_name_or_path /Qwen-7B/path/
            --vision_tower /vit-large-patch14/path/
            --freeze_vision_tower True
            --freeze_lm_model False
            --vision_select_layer  -2
            --use_im_start_end True
            --bf16 True
            --per_device_eval_batch_size 4
            --gradient_accumulation_steps 1
            --evaluation_strategy "no"
            --save_strategy "steps"
            --save_steps 5000
            --save_total_limit 1
            --weight_decay 0.
            --warmup_ratio 0.03
            --lr_scheduler_type "cosine"
            --logging_steps 1 --tf32 True
            --model_max_length 4096
            --gradient_checkpointing True
            --dataloader_num_workers 4
            --report_to none
            --per_device_train_batch_size 4
            --num_train_epochs 1
            --learning_rate 5e-5
            --datasets  data_name1+data_name2+data_name3
            --output_dir /path/to/output/

For Vary-tiny

deepspeed   Vary/train/train_opt.py  --deepspeed /Vary/zero_config/zero2.json
            --model_name_or_path /opt125m/path/
            --conversation_version opt
            --freeze_vision_tower False
            --freeze_lm_model False
            --use_im_start_end True
            --bf16 True
            --per_device_eval_batch_size 4
            --gradient_accumulation_steps 1
            --evaluation_strategy "no"
            --save_strategy "steps"
            --save_steps 5000
            --save_total_limit 1
            --weight_decay 0.
            --warmup_ratio 0.03
            --lr_scheduler_type "cosine"
            --logging_steps 1 --tf32 True
            --model_max_length 4096
            --gradient_checkpointing True
            --dataloader_num_workers 4
            --report_to none
            --per_device_train_batch_size 16
            --num_train_epochs 1
            --learning_rate 5e-5
            --datasets  data_name1+data_name2+data_name3
            --output_dir /path/to/output/

Contact

If you have any questions related to the code or the paper, feel free to email (weihaoran18@mails.ucas.ac.cn).

Acknowledgement

LLaVA: the codebase we built upon!
Qwen: the LLM base model of Vary, which is good at both English and Chinese!

Citation

If you find our work useful in your research, please consider citing Vary:

@article{wei2023vary,
  title={Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models},
  author={Wei, Haoran and Kong, Lingyu and Chen, Jinyue and Zhao, Liang and Ge, Zheng and Yang, Jinrong and Sun, Jianjian and Han, Chunrui and Zhang, Xiangyu},
  journal={arXiv preprint arXiv:2312.06109},
  year={2023}
}

cysin / Vary