Skip to content
This repository has been archived by the owner on Jun 2, 2022. It is now read-only.

waifu2x-converter-cpp ... -j ? #254

Open
idanka opened this issue Oct 13, 2020 · 1 comment
Open

waifu2x-converter-cpp ... -j ? #254

idanka opened this issue Oct 13, 2020 · 1 comment

Comments

@idanka
Copy link

idanka commented Oct 13, 2020

My question would be, what is the optimal value of "-j"?
waifu2x-converter-cpp ... -j xy?

i have i5-4570 (4 Core, 4 Thread) and nvidia 1650, 16 GB RAM
I use Ubuntu Linux 20.04
What to consider?

conversion is currently running (waifu2x-converter-cpp --scale-ratio 2.667 --noise-level 2 -m noise-scale -i /img -o /img2x --model-dir .waifu2x --processor 0 -z -j 4)

  • GPU max73%
  • GPU VRAM 442 MB
  • CPU all 10-20%
  • RAM 4.41GB

Thx!

@idanka
Copy link
Author

idanka commented Oct 28, 2020

Lol:) test 31 image (--processor 0)
gpu or cuda used, not authoritative

-- -  Done, took: 2m 26.313s total, ETA: 0.000s, file: 4.714s avg: 4.720s
-j 4  Done, took: 2m 26.342s total, ETA: 0.000s, file: 4.738s avg: 4.721s
-j 8  Done, took: 2m 26.492s total, ETA: 0.000s, file: 4.715s avg: 4.726s
-j 16 Done, took: 2m 26.583s total, ETA: 0.000s, file: 4.732s avg: 4.728s
-j 64 Done, took: 2m 26.528s total, ETA: 0.000s, file: 4.729s avg: 4.727s

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant