{"id":7089,"date":"2022-10-18T16:49:51","date_gmt":"2022-10-18T23:49:51","guid":{"rendered":"https:\/\/mattfife.com\/?p=7089"},"modified":"2023-02-09T11:41:49","modified_gmt":"2023-02-09T18:41:49","slug":"installing-stable-diffusion-1-4","status":"publish","type":"post","link":"https:\/\/mattfife.com\/?p=7089","title":{"rendered":"Installing Stable-diffusion 1.4"},"content":{"rendered":"\n

Time to play with some AI generated art! <\/p>\n\n\n\n

Here’s some great instructions on how to install the older stable diffusion 1.4:
https:\/\/www.howtogeek.com\/830179\/how-to-run-stable-diffusion-on-your-pc-to-generate-ai-images\/<\/a><\/p>\n\n\n\n

8GB of VRAM or less<\/h2>\n\n\n\n

One of the first things you’ll run into is that you’re not going to be able to generate any images at 512×512 or larger if you have a graphics card with 8GB of VRAM or less. Even smaller if you only have 4GB of VRAM. The first\/easiest method is to limit the output image size. There is also an option that splits the model into 4 parts and loads each separately<\/a> (thought it will take longer), or using a more optimized\/compressed set of trained model data. <\/p>\n\n\n\n

So how do you do that if you have an older graphics card that only has 4GB or 8GB of VRAM? TingTingin <\/a>has some tips at the end of his installation video if you are using a card with 8GB of VRAM (nVidia 3070’s for example).<\/p>\n\n\n\n

Summary (at 15:45<\/a>):<\/strong> Modify your txt2img.py and add the line ‘model.half()’ after model = instantiate_from_config(config.model) in the load_model_from_config() function.<\/p>\n\n\n\n

\n