Mr DeepFakes Forums
  • New and improved dark forum theme!
  • Guests can now comment on videos on the tube.
   
Total Likes Received: 7 (0.1 per day | 0.47 percent of total 1494)
(Find All Threads Liked ForFind All Posts Liked For) Total Likes Given: 14 (0.21 per day | 0.94 percent of total 1495)
(Find All Liked ThreadsFind All Liked Posts)

Nickrus
(DeepFaker)
DeepFaker

Registration Date: 05-15-2019
Date of Birth: Not Specified
Local Time: 07-21-2019 at 05:17 PM
Status:

Nickrus's Most Liked Post
Post Subject Numbers of Likes
RE: DeepFaceLab Explained and Usage Tutorial 1
Thread Subject Forum Name
DeepFaceLab Explained and Usage Tutorial Guides
Post Message
Thank you for this great tutorial!

I believe the initial post has too much details about models and other features, a newbie like me would prefer not to face it in the beginning, scrolling 10s of screens of histograms is scary. An advise which model to use in the beginning would be more that enough. Just my humble opinion.

Regarding the description, my experience.

(10-18-2018, 05:45 AM)dpfks Wrote: You are not allowed to view links. Register or Login to view.Available as standalone with zero dependencies, ready to be used with prebult binary (CUDA, OpenCL, ffmpeg, etc.) for all windows versions.

I tried downloading it on my Windows 7 x64, AMD CPU, Nvidia 1030. A problem happened while running 4) data_src extract faces MT all GPU.bat:


Quote: File "L:\temp\DeepFaceLabCUDA9.2SSE\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 1551, in __init__
   super(Session, self).__init__(target, graph, config=config)
 File "L:\temp\DeepFaceLabCUDA9.2SSE\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 676, in __init__
   self._session = tf_session.TF_NewSessionRef(self._graph._c_graph, opts) tensorflow.python.framework.errors_impl.InternalError: cudaGetDevice() failed. Status: CUDA driver version is insufficient for CUDA runtime version

Same problem with 4) data_src extract faces S3FD best GPU.bat

Strange, right? I have a pretty recent Nvidia driver (425.25), play games etc. After receiving this error I installed CUDA Toolkit from here: You are not allowed to view links. Register or Login to view. .

After installing and rebooting it works.

Quote:Performing 1st pass...
Running on Generic GeForce GPU #0. Recommended to close all programs using this device.
Using TensorFlow backend.
  4%|#7                                      | 74/1664 [02:17<49:16,  1.86s/it]

So maybe CUDA Toolkit is sometimes needed. But it's very slow and CPU is almost fully loaded. By the way, when running 4) data_src extract faces S3FD all GPU.bat I have a better result.


Quote:Performing 1st pass... Running on Generic GeForce GPU. Recommended to close all programs using this device.
Using TensorFlow backend.
 24%|#########1                             | 392/1664 [00:27<01:30, 14.04it/s]

Nickrus's Forum Info
Joined:
05-15-2019
Last Visit:
07-08-2019, 05:22 PM
Total Posts:
55 (0.81 posts per day | 0.64 percent of total posts) Find All Posts
Total Threads:
12 (0.18 threads per day | 0.62 percent of total threads) Find All Threads
Time Spent Online:
19 Hours, 59 Minutes, 54 Seconds
Thanks/Likes:
Given: 14 | Recieved: 7
Members Referred:
0
Nickrus's Contact Details
Nickrus's Signature
Windows 7 x64, nVidia 1030