Beyond Experimental版本v2.1 FIXED (ID: 507103)

User note using other negative embedding will make the model lose quality and or make it worse, even with my own negative embedding it’s not recommended to use other negative embedding but everyone have different tastes

Model will be out for full release soon, once it’s fully released I will move on to other models, I will also be working on pony model soon to be released

I’ll also be creating Lora’s for people who want to merge Lora’s with the models they will have no trigger words if you want to merge multiple Lora’s without the model breaking you must merge the Lora first then train the merged model then train the new Lora then merge it together again and repeat the process, when you do this the old Lora’s will no longer be compatible, it’s confusing but it works and you get better results

描述:

uploaded the wrong file

训练词语: BeyondNegative