ToDoList

• project to make one env to use different deeplearn models

use gpu

# 先安装好cuda这些
# nvcc --version 看版本
# 安装torch,从https://pytorch.org/get-started/locally/这里看pip命令，我的cuda11.7用这个
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117

projects

• FastSAM

cd FastSAM
python Inference.py --model_path .\models\FastSAM-x.pt --img_path .\images\dogs.jpg  --matting True --text_prompt "the yellow dog" --output ./

• rembg

rembg i -m isnet-general-use  .\FastSAM\images\dogs.jpg .\3.png

• vocal-remover

cd vocal-remover
python inference.py --input path\to\music --gpu 0

• GFPGAN

cd GFPGAN
python inference_gfpgan.py -i inputs/whole_imgs -o results -v 1.3 -s 2

• CodeFormer

cd CodeFormer
python inference_codeformer.py -w 0.7 --input_path [file or dir path]