Some thing about CUDA

[tychey]

HI:

When I execute demo.py

I meet the bug as below:

RuntimeError: CUDA out of memory. Tried to allocate 626.00 MiB (GPU 0; 8.00 GiB total capacity; 3.71 GiB already allocated; 469.62 MiB free; 5.58 GiB reserved in total by PyTorch)

The model I choose is bigger one: ./checkpoints_G_oilpaintbrush

The net_G I choose is zou-fusion-net

Hope your answer.

Thank you.

[tychey]

I think maybe change the batch_size will help, but I didn't find it

should I increase or decrease the arg m_grid?

[jiupinjia]

Hi, sorry for the late response.
You can try using a smaller stroke number, e.g., by reducing --max_m_strokes from 500 to 300.

Also you can try our lightweight renderer:
python demo_prog.py --img_path ./test_images/apple.jpg --canvas_color 'white' --max_m_strokes 500 --max_divide 5 --renderer oilpaintbrush --renderer_checkpoint_dir checkpoints_G_oilpaintbrush_light --net_G zou-fusion-net-light
Hope it helps. Thanks!