search

qweasdd / manga-colorization-v2

121 stars 26 forks source link

How to exclude .DS_Store files #10

Open Drizzt2028 opened 9 months ago

Drizzt2028 commented 9 months ago

When running on macOSX, an error will be reported because there is a .DS_Store file in the directory.

Drizzt2028 commented 9 months ago

` name, ext = os.path.splitext(image_name) if (image_name == '.DS_Store'): os.remove(args.path,'.DS_Store')

    print("Delete .DS_Store OK")` Worked
Keiser04 commented 9 months ago

in the inference.py? i'm editing a fork to make it more visual (gradio) so i can fix it for mac users too.

arthurwolf commented 9 months ago

@Keiser04 got a link for the fork?

Keiser04 commented 9 months ago

https://github.com/Keiser04/manga-colorization-v2

copy me the error and send it to me either here or in an issue in my fork

so you save me work and I try to fix the error.

DDXDB commented 9 months ago

@Keiser04 Do you know how to train

Keiser04 commented 9 months ago

@DDXDB Yes but the model only works in v1, the v2 has encoders but idk how i can do

Keiser04 commented 9 months ago

but... idk how to use. this is my best result image

arthurwolf commented 9 months ago

is that result with the default provided model, or one you trained yourself ? anyone around here interrested on sharing efferts on training (like working on preparing a dataset together, and on training together ?)

Keiser04 commented 9 months ago

It is my model

Keiser04 commented 9 months ago

is that result with the default provided model, or one you trained yourself ? anyone around here interrested on sharing efferts on training (like working on preparing a dataset together, and on training together ?)

Look inside my repository, I have made some to facilitate the dataset. if you want I can share it with you but the results are not the desired and it takes a long time.

arthurwolf commented 8 months ago

Look inside my repository, I have made some to facilitate the dataset. if you want I can share it with you but the results are not the desired and it takes a long time.

I plan on starting a dataset of my own in a few weeks. Would you be interested in joining efforts? Put our datasets together (and maybe find other participants here too?), and also share efforts on training. The result would be a model better than what either of us can do. If you're ready to share your dataset/model, I'll come back to you when I have my own and merge the two, and see what training with the merged dataset does.

As they say, Unity is Strength.

Keiser04 commented 8 months ago

Look inside my repository, I have made some to facilitate the dataset. if you want I can share it with you but the results are not the desired and it takes a long time.

I plan on starting a dataset of my own in a few weeks. Would you be interested in joining efforts? Put our datasets together (and maybe find other participants here too?), and also share efforts on training. The result would be a model better than what either of us can do. If you're ready to share your dataset/model, I'll come back to you when I have my own and merge the two, and see what training with the merged dataset does.

As they say, Unity is Strength.

Let's see, yes, I can share my dataset with you, but I honestly don't trust the results I had.