
I know people will hate the model costs $4 (4,000 buzz points). But, for future models. I have decided to manually create prompts for every single image during the training process to better the training process. Because, some model have 30-hundreds of images that google colabs miss information on art. I think given I want to create better ai models. An to create better character models. I need to put in some extra work.
This way the ai learns more information from the images. Similar to how google captcha learns from people feeling in the gaps. Because, I will have to put in alot more work to make character specific SDXL Lora files. I think it's only reasonable that people pay a bit more buzz to help better ai models. Which is only fair.
You can only get better results when, people contribute. So, in order to make better models. The increased buzz requirements only seems natural. I am sorry if this annoys people. But, in all seriousness. $4 isn't that much I'm going to be real. I'm going to be working on more Nintendo and work harder to get character models right. So increased training time, prompts, and sessions are pretty much a must. In order to create better models.
The models will remain free to use. But I hope people understand my reasoning for jumping from $0.05-0.50 to $4 per base character model. Since, people who don't train models. May not understand why the price jump is kind of warranted. If people want higher quality character models.
描述:
训练词语: funky kong
名称: FunkyKong.safetensors
大小 (KB): 166543
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success