emikulic·3 years ago·discussThree epochs means it sees each token three times. The dataset is ~1T like you said.
emikulic·3 years ago·discussI want flying cars and moving sidewalks like I was promised by The Jetsons. :(
emikulic·3 years ago·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 years ago·discussEmad said SD cost $600,000 to train. I wonder if Midjourney also had to pay that to train from scratch.
emikulic·3 years ago·discussWhat benefits does the Huggingface diffusers(?) implementation have over A1111?
emikulic·3 years ago·discussI like to run A1111 in --api mode and write my own script to drive it over HTTP.
emikulic·3 years ago·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 years ago·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 years ago·discussI hope you will run your own locally instead of letting Midjourney block and/or mess with your prompts.
emikulic·4 years ago·discussIf edge devices includes my gamer PC then yes. For apps I'd recommend https://github.com/AUTOMATIC1111/stable-diffusion-webui