The trained Doom slayer model doesn't work that well, but it still works
Description
Using more assets to optimize is much better, but it requires more training
使用了更多素材进行优化，效果好了不少，但是需要更多训练。
FAQ
Comments (9)
I'm not sure if it's just me, but he kind of looks... plastic. Seems like the lora needs some more time to cook.
The trained Doom slayer model doesn't work that well, but it still works，i need more images to train ，can you help me?
@74120 I've wanted to make loras myself, but I'm not sure how. What kind of images do you need?
@MiniMage Doom slayer's images ,I need it to upgrade the model.
@74120 Any specific perspectives or something? You seem to be generating a Doom: Eternal doomslayer, so wouldn't searching doom: eternal work? I did a search on DuckDuckGo, and I got pages upon pages of pictures. But I don't know what you have already. Besides, how would I send you the pictures? Version 1 looked more like pre-Doom:Eternal doomslayers. I personally would've shared a lot of reference picture from Doom(2016) as I prefer the look of that one myself. Again, it depends entirely on what you need.
@MiniMage I get training images from other websites, and if there are more images, the training will be better, and if you want to send me images, you can use Twitter.my id is@WzrUOs .If you can give me enough images for two versions, I'm going to train two models separately, thanks
dude why can i find ur lora when i look it up on civit ai