>What is this?A local open weights music generator, like Suno and Udio.>Original repo (includes lora training)https://github.com/ace-step/ACE-Step-1.5>Comfyui guide (but use the SFT model instead of Turbo, CFG=1 and 50 steps)https://docs.comfy.org/tutorials/audio/ace-step/ace-step-v1-5>Suno-like UIshttps://github.com/fspecii/ace-step-uihttps://github.com/roblaughter/ace-step-studio>Cover and Edit modeshttps://github.com/ryanontheinside/ComfyUI_RyanOnTheInside/tree/main/examples/ace1.5>Cover and reference song tutorialhttps://www.youtube.com/watch?v=sv4pNrjRh7sShare your gens and lora results.Keywords: music gen, local model, song gen, suno, udio, acestep, ace step, lmg,ldg, dmp
Britney Spears lora:https://voca.ro/16blE7la2Ff8https://voca.ro/1eLHswbE9ZHKhttps://voca.ro/1me4VBIkzHfKTo that one anon that is claiming "lora training on SFT doesn't work": this is a Lora trained on SFT =)
based thread! i don't use ACE Step but the best wisheS!
>>108164777Nakadashee AceStep-chan
>>108165006it is...ok, can hear her singing style from time to time.my internets is cutting out for half a day and more lately since i live in a third world country (australia), so yesterday i had time to sit and test default settings as per developer instruction, results were meh via gradio.must test comfy nodes since i got better results than via gradio interface.one note, manual captions, removing redundant stuff like too many attributes llm gives (in case it does detect correct instruments), help quite a bit.enya anon done it really well via overfit, and if he sees this post;what was your overfit setting?high lr low rank low small-medium dataset?
>>108165607>enya anon done it really well via overfit, and if he sees this post;>what was your overfit setting?>high lr low rank low small-medium dataset?I used the default 0.0003 LR at 800~1000 epochs (I can't remember where I stopped), my dataset consisted of 24 songs
>>108165764As of the rank, I used rank 128 I think, it's the maximum my card supports without going OOM
>>108165776>>108165764ty, will try it. i used 12 songs, same genre and different bands, results are meh and sometimes ok.comfy gens are better, if i crank lora strength to 2 there is that fm radio but super high can+static noise, yet it does replicate training set song at around 80% of the content.
>>108166294>fm radioAM -.- radio
>>108166294>comfy gens are better, if i crank lora strength to 2 there is that fm radio but super high can+static noiseI said that in last thread and I am going t say again, you are probably undertraining your models. Use a high enough LR, train longer, and use a high rank.
>>108166359Also DO NOT USE THE LLM.The LLM tends to weaken the Lora effect, sometimes it even changes the voice/singing style
does anyone have access to the suite Sony has produced to compare?
>>108166444>suite Sony has producedWhat even that is?
>>108165776In my experience 128 has been good. I think people need to cook their LoRAs a lot longer though. At 1500 epochs 2000 or more is probably better.>>108166395Yeah do not use the LLM if you are using a LoRA. It straight up makes it own song and the LoRA is just antagonistic to it.
