NokiMo
Furkan Gözükara
Furkan Gözükara

patreon


1 Click Download 160 Very Best Stable Diffusion 1.5 Based (SD 1.5) Models

Join discord and tell me your discord username to get a special rank : SECourses Discord

Patreon exclusive posts index

Login or Register RunPod : https://runpod.io?ref=1aka98lq

23 January 2024 Update

18 January 2024 Update:

I have spent over 1 day to code a special python script that will download the very best 160 models from CivitAI and Hugging Face.

Total size is 409 GB.

You can edit the download_top_160_SD_1_5_models .py file and disable download of any model you want.

Why I prepared this?

Currently I am doing testing on each model. Once I find the most realistic and advanced one, hopefully I will do academic experiment on that model to find out very best DreamBooth training configuration as I did for SDXL.

On Windows download_top_160_SD_1_5_models. bat file and download_top_160_SD_1_5_models. py file and double click bat file to start downloading.

If any error occurs during download, it will move to the next model. After entire download ended, run the script again and it will download only error occurred models. Such errors occurs since this is a network based script.

For runpod just upload download_top_160_SD_1_5_models .py file into workspace folder and run as below

When you run the script, it will ask you how many threads to download. 1 works best but based on your network speed, you can enter more such as 5.

You can also edit py file and set the download folder as you wish.

The files will be downloaded into the stable-diffusion-webui/models/Stable-diffusion folder by default wherever the script is run. You can edit it.  

Automatic Installers for Automatic1111 SD Web UI

Comparison x/y/z Plot Results PNG Images

Click To Download More Comparisons (10 X/Y/Z plots - 1.1 GB)

  1. Image Of A Man - 768x768 - full resolution - 1.1 GB - right click save as
  2. Image Of A Man - 768x768 - 50% resolution - 300 MB - right click save as
  3. Image Of A Man - 768x768 - 25% resolution - 77 MB - right click save as
  4. Photo Of A Man - 768x768 - full resolution - 1.1 GB - right click save as
  5. Photo Of A Man - 768x768 - 50% resolution - 300 MB - right click save as
  6. Photo Of A Man - 768x768 - 25% resolution - 77 MB - right click save as
  7. Picture Of A Man - 768x768 - full resolution - 1.2 GB - right click save as
  8. Picture Of A Man - 768x768 - 50% resolution - 300 MB - right click save as
  9. Picture Of A Man - 768x768 - 25% resolution - 77 MB - right click save as
  10. Photo Of A Man - 768x1024 - full resolution - 1.5 GB - right click save as
  11. Photo Of A Man - 768x1024 - 50% resolution - 378 MB - right click save as
  12. Photo Of A Man - 768x1024 - 25% resolution - 100 MB - right click save as
  13. Photo Of A Man - 768x1024 - full resolution - 2 GB - right click save as
  14. Photo Of A Man - 768x1024 - 50% resolution - 500 MB - right click save as
  15. Photo Of A Man - 768x1024 - 25% resolution - 128 MB - right click save as

1 Click Download 160 Very Best Stable Diffusion 1.5 Based (SD 1.5) Models

Comments

yes it downloads all models as .safetensors model

Furkan Gözükara

Is it possible to download the files in the safetensor model? Because I don't have Stable Diffusion installed on the machine, I use it on websites...

Gpr0mpt

yes i am planning hopefully

Furkan Gözükara

Thank you very much for this! Do you plan to do the same for SDXL models? :)

OctonionPrime

Oh I see. yes that can be the reason. My auto installers are expecting CUDA and NVIDIA drivers to be installed.

Furkan Gözükara

I think I might have found the issue. I'm using an EC2 instance on AWS and by default it didn't have the nvidia drivers installed on it. I'll try installing them and then see how that goes.

Robert White

hello. what is your GPU? if AMD you need to add --skip-torch-cuda-test and some other parameters.

Furkan Gözükara

This is for the auto1111 install

Robert White

I've got your automatic downloader but I'm running into this error when running it. "RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check"

Robert White

thanks a lot for sharing. i checked a lot of models but must have missed

Furkan Gözükara

Oo nice, but you have missed one really interesting model :-) https://civitai.com/models/148165/puredarkness Just kidding, but you have to check it out - i'v Finetuned it with almost 1000 "cheery-awesome-picked" images that i created in Illuminati Diffusion... I think it is kinda unique model, you can do a lot of awesome imgs with it. I find out it is really great for merging with other models - so you can try to merge it with your Dreambooth model trained on your face etc...

Nenad Kuzmanovic


Related Creators