Can you please release a pretrained model?

[seven7777777]

Hi, thanks for your great job. I want to know if it is possible for you to release a pretrained model. thank so much!

[shangdibufashi]

+1

[maria-larchenko]

Hi, @seven7777777 @shangdibufashi!
Thank you for your interest, we do plan to release a pretrained model, but it will take another week or two. We additionally plan to publish code for our reproduction of Neural Preset (https://github.com/ZHKKKe/NeuralPreset). However without a collection of LUT filters it doesn't seem to work as intended by the authors.

[seven7777777]

Hi, @seven7777777 @shangdibufashi! Thank you for your interest, we do plan to release a pretrained model, but it will take another week or two. We additionally plan to publish code for our reproduction of Neural Preset (https://github.com/ZHKKKe/NeuralPreset). However without a collection of LUT filters it doesn't seem to work as intended by the authors.
@maria-larchenko
wow, that's really cool!

[xiaomeixivvr]

嗨， ！感谢您的关注，我们确实计划发布预训练模型，但还需要一两周的时间。我们还计划发布用于 Neural Preset （https://github.com/ZHKKKe/NeuralPreset） 复制的代码。但是，如果没有 LUT 过滤器的集合，它似乎无法按照作者的预期工作。

May I ask if you plan to release the pre-trained model yet? This is great work!

[shangdibufashi]

Hi, @seven7777777 @shangdibufashi! Thank you for your interest, we do plan to release a pretrained model, but it will take another week or two. We additionally plan to publish code for our reproduction of Neural Preset (https://github.com/ZHKKKe/NeuralPreset). However without a collection of LUT filters it doesn't seem to work as intended by the authors.

i have penalty lut files 5000+, please contact me if you need it

[Yangkai-Wei]

@shangdibufashi I am reproducing the Neural Preset and also encountered the problem of poor results without LUT files. Could you please share your LUT files with me ? Thank you very much.

[maria-larchenko]

@seven7777777 @shangdibufashi @xiaomeixivvr

please check the hugginface link
https://huggingface.co/MariaLarchenko/modflows_color_encoder/tree/main

There are three checkpoints, B6 (EfficientNet B6) is the main model and B0 (EfficientNet B0) was used for ablation study.
To use these models, refer to the demo notebook
https://github.com/maria-larchenko/modflows/blob/main/ModFlows_demo.ipynb