I use 3090 with 24gb Vram but I can only train fp8 lora. I use mass compute to train fully fp16 lora
I use 3090 with 24gb Vram but I can only train fp8 lora. I use mass compute to train fully fp16 lora with the cost of about 4 usd
Wvery cheap
VThat is per hour?
Wno, total cost
OMassed Compute - from my experience has been great, far better than runpod. The SECourses pricing on the A6000's is great.
Wper hour is only 1.25 usd
Wwhen training with my local pc, I need 11 hours of 3090 run full throttle, my poor pc fan roar the hell of of my room and the temperature turn up a lot
Wso 4 usd saving my trouble and time
W11 hours of 3090 for fp8 lora
Wno thank you
DI'm training this model now, Works much better for training, you can train multiple people on one lora without concept mixing, it behaves very much like sdxl, much more flexible
VYeah the pc lights stayed on and it looks like a disco ballroom.
WI never see any loras in SDXL have multiple concept, Flux maybe the first model can do it
DI trained loras on sdxl with more than 20 people no problem
WSince SDXL have instance ID, so I use it for more convenience if I need people lora anyway
VBest gpu prices are from here - https://www.autodl.com/home but I don't know how to run them. Learned about them via an article discussing 48gb 4090s.
AutoDL为您提供专业的GPU租用服务，秒级计费、稳定好用，高规格机房，7x24小时服务。更有算法复现社区，一键复现算法。
Wthis really new to me
OIs it possible to create a finetune, then extract the LORA? I've had good luck with that for single subjects, and sometimes it's also nice to have a checkpoint as well.
Dflux-dev can train only one subject, but with flux-dev-de-distill I can learn many concepts
WMost of us dont know chinese, oh well
V@wardensc2 I just use google translate to change the text.
Dyes but is too slow... now I'm training on my 3090ti, may be then I will try on mass compute
OI don't think I've ever tried to train a finetune with more than one subject
Wbut when you access to their server, their GUI also chinese, what will you do = )))
V@wardensc2 Click translate in the browser, lol.
V
Dyes but the quality is much worse... now I'm full on flux-dev-de-distill
WAlso @Dr. Furkan Gözükara already set up kohya_ss for quick install on Mass Compute, I just need 1 console code for it to run everything
Wno this is only the beginning
Wsimilar to masscompute, you have to upload your image training data and the checkpoint you want to train, vae, clip and install Kohya as well
Wit similar to our local PC but better hardware, it's not one click training.
WNo I guarantee when it come to people image, install ID is much better than lora if you know how to use it properly
WI'm not sure I can upload AI image of my instant ID for you to see on discord
WI make Mbappe Instant ID, not sure SDXL lora can beat it
DI know what you mean I try it, but the quality over all is much better on flux
DSDXL is good but not as good
Wof course I dont compare with Flux but SDXL lora can't beat Instant ID
V@wardensc2 I understand. Im just pointing at the price difference.
NVIDIA L20 / 48GB single precision 59.35 TFLOPS / semi precision 119.5 Tensor TFLOPS ￥ 4.19 / hour
USD is 0.60 an hour.
NVIDIA L20 / 48GB single precision 59.35 TFLOPS / semi precision 119.5 Tensor TFLOPS ￥ 4.19 / hour
USD is 0.60 an hour.
VNVIDIA L40 / 48GB single precision 90.5 TFLOPS / semi precision 181.05 Tensor TFLOPS ￥ 4.82 / hour
$0.69 USD
$0.69 USD
DIDK with a well trained lora you get perfect likeness and more flexibility
Wyes I know but when you using it, you can't translate everything if the operating system and other support software are also chinese
VI see, your saying the whole OS would be in chinese.
Dbut its ok if works for your purpose
VMaybe Dr. Can team up with them and have them make an option for flux training preconfigured.
Wnormally I just gen a image with style and pose I want then apply Instant ID on it. SDXL lora can't flexible than this
Wsomething like this
Dlooks very good
Wthis is SDXL not Flux
Ois instant id coming to flux?
Wmore fantasy ways
WI haven't heard it yet