Does the current <11gb file limit problem mean its currently impossible to train a FLux dev lora and use in my own workflow?

[Dawgmastah]

TItle
Does the current <11gb file limit problem mean its currently impossible to train a FLux dev lora and use in my own workflow?

Is there another way to perhaps train a lora and use it from the
https://replicate.com/fofr/any-comfyui-workflow/ model? I have been trying it to no success