这项目和 Flux 有半毛钱关系吗？

[wikeeyang]

感觉这就是个蹭 Flux 流量的项目，白花了我半天时间来配置环境和测试，硬代码一大堆，都需要自己手工改。
不是说尽快修改硬代码吗？自己弄的代码，自己都搞不掂？
我看 Iuuses 里，各种问题全是各路大神在帮你搞掂。。。
这半吊子开源来干嘛？

[bringrainfire]

If what I translated is correct, then that's a very crappy attitude to have against someone that's writing something you want to use for free. Take a step back and chill out.

[wikeeyang]

@bringrainfire I appreciate the spirit of open source and acknowledge the contributions of the authors, but this is technology and science, which requires a rigorous attitude. The purpose of open source is to share ideas, discuss together, and make progress together.
The new approach of this project is completely commendable, but for some unknown reason, the original author did not demonstrate the effects brought by this new approach (Sample), did not compare the effects of similar products, and did not improve the hard code used during training, which caused everyone who tried and tested the source code to waste a lot of time adjusting and figuring out what the effects were good about? A person can easily do better, share better, but waste a lot of people's time unnecessarily. Is this attitude worth encouraging at the technical and academic research levels?
I spent half a day flipping through all the contributors' code in Issues and have tested it, but the results were not particularly impressive. Flux in the T2I field is a stunning project with a very good model foundation and generalization ability, taking T2I to a new height. What are the highlights of this project?

[wikeeyang]

I refer to @paibenjam, @curved, and "https://huggingface.co/spaces/flosstradamus/FluxMusicGUI", after modifying a lot of code themselves to complete the test, many enthusiastic people have contributed so much good code and ideas, Why didn't the original author update the code?

[wikeeyang]

This is the had cleaned code???

[charfeng1]

所以这个现在是用不了的吗