StyleCLIP GUI Windows Installation Tutorial

My Copy of StyleCLIP GitHub

Colab Tutorial

Notes:
---------------
## Anaconda3 Download:

## system variable CUDA 10.2 move to the very top

## CUDA 10.2 Toolkit download:

conda create -n StyleCLIP python=3.7
conda activate StyleCLIP
conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=10.2 -c pytorch

## for 30 series
conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=11.0 -c pytorch

pip install ftfy regex tqdm gdown
pip install opencv-python
pip install tensorflow-gpu==1.14
pip install matplotlib

##cd to the right file directory

cd global
# builds W.npy
# builds S

# input image inside data/ffhq
# corresponding latent space file in

---------------

My main channel where I introduce the latest fascinating AI tools

Related video for info/explanation

[Seacage]

Hi, thank you for the tutorial. Could you also explain also how to uninstall everything afterwards please?

[LordLethris]

Has anyone done a CPU version of this?
Because not the whole world revolves around nVidia GPUs.
I'm very confident that my CPU is capable, even if it will take a bit longer. Tensorflow already supports a CPU only environment and so does pytorch :)

[fate]

I don't have this option "Anaconda Prompt". What shoud I do?

[Igrom_Gamedev]

I got hung up at the point where I need to get latents.pt. In coloboration, either there are no free GPUs, or it gives errors in the first item. Is there any way around this?

[detroxlp]

dont work...

It said:

ImportError: Could not find 'cudart64_100.dll'.

Then I downloaded it manually from the internet and run code again. And it say:


So I sign up to download cudnn and now it says: Failed to load the native TensorFlow runtime.

Wasted 4 hours of my time...I did *everything* exactly how in the video!

[Remo-media]

Wow this is so broken.. I think you need to redo this Tut or delete
Many many errors been at this for 12 hours
..fatal error C1083: Cannot open include file: 'windows.h': No such file or directory

fused_bias_act.cu

[odio]

Getting an error message whenever I run "python GetCode.py --dataset_name ffhq --code_type w." The input I get is as follows, and yes, I do have python installed. :P

Traceback (most recent call last):
File "C:\Users\Ande\anaconda3\envs\StyleCLIP\lib\site-packages\tensorflow\python\platform\self_check.py", line 75, in preload_check

File "C:\Users\Ande\anaconda3\envs\StyleCLIP\lib\ctypes\__init__.py", line 364, in __init__
self._handle = _dlopen(self._name, mode)
OSError: [WinError 126] The specified module could not be found

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "GetCode.py", line 7, in <module>
from dnnlib import tflib
File "S:\2\StyleCLIP\global\dnnlib\tflib\__init__.py", line 9, in <module>
from . import autosummary
File "S:\2\StyleCLIP\global\dnnlib\tflib\autosummary.py", line 28, in <module>
import tensorflow as tf
File "C:\Users\Ande\anaconda3\envs\StyleCLIP\lib\site-packages\tensorflow\__init__.py", line 28, in <module>
from tensorflow.python import pywrap_tensorflow # pylint: disable=unused-import
File "C:\Users\Ande\anaconda3\envs\StyleCLIP\lib\site-packages\tensorflow\python\__init__.py", line 49, in <module>
from tensorflow.python import pywrap_tensorflow
File "C:\Users\Ande\anaconda3\envs\StyleCLIP\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 30, in <module>
self_check.preload_check()
File "C:\Users\Ande\anaconda3\envs\StyleCLIP\lib\site-packages\tensorflow\python\platform\self_check.py", line 82, in preload_check
% (build_info.cudart_dll_name,

[core]

thank you for your tutorials. i just love the things people do with ai, but without your tutorials i would probably never be abe to try any of the tools

[CutterboyBK]

when i use "conda create -n StyleCLIP python=3.7" i get an error message that says "'conda' is not recognized as an internal or external command, operable program or batch file"

[insta_]

I get stuck with ModuleNotFoundError: No module named 'tensorflow'

[PizzaBoy]

are you still able to run the e4e completion in the colab? I just get a UnpicklingError on the line: ckpt = torch.load(model_path, map_location='cpu')

[xoshaow]

e4e completion errors every time. ckpt = torch.load(model_path, map_location='cpu'), aligned_image = align_face(filepath=image_path, predictor=predictor). any suggestions on how to get this fixed? thanks!

[Shock_Treatment]

This definitely needs to be updated or something. It doesn't work anymore.

[timothyalper]

I happen to have a 30series GPU. In the video it says untested do you think that it could still work? More specifically is there any chance of it harming my computer?

[candy-cxfu]

I would like to ask if the e4e colab cannot be used now?

[tcla]

ERROR: Cannot find command 'git' - do you have 'git' installed and in your PATH?

[gyrro]

I should've figured this wouldn't work on an AMD system judging by the fact you need to install the CUDA toolkit. Regardless I attempted anyway and ran in to issues after running the python commands after doing "cd global", none of them would work as the nvcuda.dll is not present in System32 since my system does not have an nVidia card. If anyone here is running an AMD GPU in there system, do not attempt this. It will not work, use my rather silly error as a lesson.

[iLEZ]

When running the command that builds w, I get an error message after the text you get, it starts with it failing to run cuBLAS routine: etc etc, then failures to launch other things. "Internal Blas GEMM launch failed" for example.

[_imV]

I need to turn 2 images into realistic. can i just send them to you? i dont know how to install any of this ^^

[levenin]

Please make Real-Esrgan google COLAB TUTORIAL...PLEASE