Wow! Cool creation. Thanks for including my Flaming Inferno DSv8 LoRA in the mix. Unlike the Flux LoRAs, when blending in 1.5 models, you often need to adjust the wt of the models within the token code to get them to play together nicely. I can only see what I presume is the start of your prompt in the "title" area and am going from that. Also, if you didn't use one, a negative prompt (even the standard default is quite effective) usually clears up the composition a lot as well. Every week I host a "blend it battle" challenge featuring 2 specified 1.5 models to blend together (and add whatever other models or prompting you like), so maybe you'd be interested? Let me know if you would like to be notified when the next challenge is posted tomorrow.
Wow! Cool creation. Thanks for including my Flaming Inferno DSv8 LoRA in the mix. Unlike the Flux LoRAs, when blending in 1.5 models, you often need to adjust the wt of the models within the token code to get them to play together nicely. I can only see what I presume is the start of your prompt in the "title" area and am going from that. Also, if you didn't use one, a negative prompt (even the standard default is quite effective) usually clears up the composition a lot as well. Every week I host a "blend it battle" challenge featuring 2 specified 1.5 models to blend together (and add whatever other models or prompting you like), so maybe you'd be interested? Let me know if you would like to be notified when the next challenge is posted tomorrow.