emikulic·3 ปีที่แล้ว·discussThree epochs means it sees each token three times. The dataset is ~1T like you said.
emikulic·3 ปีที่แล้ว·discussI want flying cars and moving sidewalks like I was promised by The Jetsons. :(
emikulic·3 ปีที่แล้ว·discussTry https://huggingface.co/ehartford/WizardLM-7B-Uncensored and related models. They're not even trained on smut, just the neutered responses were removed before the RLHF stage (IIUC)
emikulic·3 ปีที่แล้ว·discussEmad said SD cost $600,000 to train. I wonder if Midjourney also had to pay that to train from scratch.
emikulic·3 ปีที่แล้ว·discussWhat benefits does the Huggingface diffusers(?) implementation have over A1111?
emikulic·3 ปีที่แล้ว·discussI like to run A1111 in --api mode and write my own script to drive it over HTTP.
emikulic·3 ปีที่แล้ว·discussLast time I looked into this, the answer was "because huggingface transformers and torch.load are written to do it this way"You could absolutely do something streaming, or mmap the weights instead of loading them into system RAM. Just the default interfaces don't.
emikulic·3 ปีที่แล้ว·discussThis used to be the case but nowadays you can just pip install e.g. https://libraries.io/pypi/nvidia-cuda-runtime-cu11
emikulic·4 ปีที่แล้ว·discussI hope you will run your own locally instead of letting Midjourney block and/or mess with your prompts.
emikulic·4 ปีที่แล้ว·discussIf edge devices includes my gamer PC then yes. For apps I'd recommend https://github.com/AUTOMATIC1111/stable-diffusion-webui
emikulic·4 ปีที่แล้ว·discussYou have to create an account and agree to some stuff and be logged in but you're correct that you don't have to pay.
emikulic·4 ปีที่แล้ว·discussSame. I've found the biggest problem was SNAT rewriting the (source) port number. netfilter, by default, doesn't do this. pf does but you can configure it not to.