Search found 37 matches
- Fri Jun 19, 2020 4:25 am
- Forum: Hardware
- Topic: VRAM Usage Limits to 80%
- Replies: 12
- Views: 7464
Re: VRAM Usage Limits to 80%
I'm not so sure it's a Windows issue. I noticed that your VRAM is reporting 10312 of 11Gb used, which would be an almost identical amount to my loss (3357 of 4Gb). Even tho you're on Unix vs Windows? If that's the case then people would get about 7325 with 8Gb of VRAM and 23325 with 24Gb. So with al...
- Thu Jun 18, 2020 6:04 am
- Forum: Hardware
- Topic: VRAM Usage Limits to 80%
- Replies: 12
- Views: 7464
Re: VRAM Usage Limits to 80%
Yeah, I found that page also. But all it does is list the "compute capability" for their various NVidia products. It never says exactly what it is or how the compute capability is computed. It doesn't really matter - it's basically a marketing number. I'll stick with clock speed, number of...
- Wed Jun 17, 2020 4:55 am
- Forum: Hardware
- Topic: VRAM Usage Limits to 80%
- Replies: 12
- Views: 7464
Re: VRAM Usage Limits to 80%
Here is some additional info. In my research, people kept referring to a program called nvidia-smi.exe to get the info as Windows sees it. At first I couldn't find it in places it was supposed to be, and apparantly neither can anyone else. It should be in the NVIDIA Corporation\NVSMI directory, whic...
- Wed Jun 17, 2020 4:02 am
- Forum: Hardware
- Topic: VRAM Usage Limits to 80%
- Replies: 12
- Views: 7464
Re: VRAM Usage Limits to 80%
BTW, what is "compute capability"?
- Wed Jun 17, 2020 3:59 am
- Forum: Hardware
- Topic: VRAM Usage Limits to 80%
- Replies: 12
- Views: 7464
Re: VRAM Usage Limits to 80%
Here is what I get: 2020-06-16 17:37:41.551462: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1304] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 3001 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1050, pci bus id: 0000:3f:00.0, compute capability: 6.1)...
- Wed Jun 17, 2020 3:01 am
- Forum: Training Discussion
- Topic: The faces of the figures in the second and third columns of the preview image are white frames
- Replies: 5
- Views: 2828
Re: The faces of the figures in the second and third columns of the preview image are white frames
You should see changes even in the first preview (100 iterations). It won't lookd recognizable as a face at that point, but you should see rapid improvement over the first few thousand iterations. In addition, the error graph should fall very rapidly at first. In addition to possible corruption, I w...
- Wed Jun 17, 2020 2:45 am
- Forum: Extract Support
- Topic: Error when extracting on Mac using AMD GPU
- Replies: 3
- Views: 3025
Re: Error when extracting on Mac using AMD GPU
I could be wrong on this, but I think FaceSwap currently only works with NVidia GPU's. The only other option is CPU (Intel or AMD okay there).
- Wed Jun 17, 2020 2:32 am
- Forum: Training Discussion
- Topic: DLIGHT Diverged....
- Replies: 4
- Views: 2632
Re: DLIGHT Diverged....
I was going to add that you might want to double check where your A set is pointing. I got a very similar looking problem when I accidentially changed to set things up on a different project, and then didn't change one of my sets back properly. But then I noticed that the error is increasing, and I ...
- Fri Jun 12, 2020 9:24 am
- Forum: Hardware
- Topic: VRAM Usage Limits to 80%
- Replies: 12
- Views: 7464
Re: VRAM Usage Limits to 80%
I can't see any memory info at all using Verbose, but I'll try Debug later. On the console it stops trying to get less at about 3.2Gb, but I'll switch logs to Debug to see if I can get specifics. I don't think it's a Windows reserve because that would show up in the total VRAM allocation (i.e. Windo...
- Fri Jun 12, 2020 8:56 am
- Forum: Training Discussion
- Topic: At what point do faces start to 'pop' ?
- Replies: 2
- Views: 3257
Re: At what point do faces start to 'pop' ?
If you're at 300K and things are still fuzzy, you probably need to do some better curating. The model you're using could be it, but more likely it's just quality. I'm very restricted in the models I can use because of my VRAM (4Gb), but DFL-SAE 64-bit works very well. It is VERY important to get goo...
- Fri Jun 12, 2020 7:25 am
- Forum: Hardware
- Topic: VRAM Usage Limits to 80%
- Replies: 12
- Views: 7464
VRAM Usage Limits to 80%
I have a 4Gb (Nvidia 1050) GPU, but only 3.2Gb (exactly 80%) ever gets allocated, according to Windows performance manager. It doesn't seem to matter what it's doing (extract, training, sorts, convert), the maximum ever allocated is 80%. This represents a large performance decrease in an already lim...
- Tue Jun 09, 2020 1:36 am
- Forum: Training Discussion
- Topic: High level model algorythm descriptions
- Replies: 4
- Views: 2658
Re: High level model algorythm descriptions
That's kind of what I've been trying to do. However, my programming peak was back in the late 90's with Visual Basic, SQL databases and event driven interfaces. All of the underlying technology - python, tensorflow, keras, programming environments in general are new to me. I wrestled with Anaconda a...
- Tue Jun 09, 2020 1:20 am
- Forum: Training Discussion
- Topic: High level model algorythm descriptions
- Replies: 4
- Views: 2658
Re: High level model algorythm descriptions
I'll give that a try. I understand most of the operations involved, but not how the models differ from each other in their high level game plans. Excuse me while I duck down this rabbit hole....
- Sat Jun 06, 2020 11:50 pm
- Forum: Training Discussion
- Topic: High level model algorythm descriptions
- Replies: 4
- Views: 2658
High level model algorythm descriptions
I've been trying to file some place that has good, high-level descriptions of the various algorythms used by each model. I've seen some generalized information, such as the descriptions in the Training Guide. Those are good, but a bit too broad for me. I've also looked through the Python code, but i...
- Sat Jun 06, 2020 11:35 pm
- Forum: Cloud Discussion
- Topic: how to get the preview?
- Replies: 2
- Views: 18211
Re: how to get the preview?
Although the timelapse images may look cool, they are limited to the 14 images selected. The previews, on the other hand, eventually contain all of the images, so are more valuable than the timelapse in judging actual progress. To collect these previews I made a batch file (at bottom) that looks for...
- Sat Jun 06, 2020 9:02 pm
- Forum: Extract Discussion
- Topic: Flipped face on extreme profiles
- Replies: 8
- Views: 5475
Re: Flipped face..
I have had this problem as well, and not just on full profiles. It happens even in clear 3/4 views sometimes. You'll see seemingly identical face detected, and then for some reason some will flip upside down even with the face registered perfectly. If you need the flipped faces you CAN get them to f...
- Sat Jun 06, 2020 8:20 pm
- Forum: Convert Support
- Topic: Convert Function - Model could not be found ERROR
- Replies: 4
- Views: 3219
Re: Convert Function - Model could not be found ERROR
Just a note on this - the field can contain the correct file path and name, but you need to use the browse and select to get the file. Many times (such as switching from initial to final training sets) I have tried to just change the path or filename in the box. However, when the model runs, you get...
- Thu May 14, 2020 12:03 am
- Forum: General Discussion
- Topic: So I got probably a stupid question
- Replies: 3
- Views: 4059
Re: So I got probably a stupid question
Hmmmm. Well, they do have that for Keanau, Nicholas and The Donald, but honestly attempting to grok what makes a person's face unique during the culling is part of the artistic fun.
- Sun May 10, 2020 1:08 am
- Forum: General Discussion
- Topic: So I got probably a stupid question
- Replies: 3
- Views: 4059
Re: So I got probably a stupid question
I'm a bit confused by your use of the word "scenes". FaceSwap doesn't look at anything in a scene except for the face. The extraction pulls out only faces, training trains only faces, and swapping only glues the face from the model on top of the face in the original video. So as far as fac...
- Sun May 10, 2020 12:02 am
- Forum: Training Discussion
- Topic: Prioritize Eyes?
- Replies: 5
- Views: 5970
Re: Prioritize Eyes?
I think the problem you note is not really a problem. When a model is swapped, only the parts of the model inside of the frame are output, so anything falling outside (even half an eyeball) are okay. Ditto with training - anything outside of the frame is ignored. So the program doesn't attempt to ei...