Search found 8 matches

by elon
Thu Jan 07, 2021 10:06 pm
Forum: Installation Support
Topic: [Guide] Using Faceswap on Nvidia RTX 30xx cards
Replies: 124
Views: 201047

Re: [Guide] Using Faceswap on Nvidia RTX 30xx cards

torzdf wrote: Thu Jan 07, 2021 8:02 pm
abigflea wrote: Wed Jan 06, 2021 4:26 pm

write permissions error on second attempt after reinstalling miniconda.

That's between you and Bill Gates to resolve ;)

I solved all my 3090 problem now.

Very Satisfied.

it is Same or higher speed than 2080ti x 3way

by elon
Thu Jan 07, 2021 11:31 am
Forum: Hardware
Topic: 3x 2080s or 1x 3090
Replies: 1
Views: 10470

3x 2080s or 1x 3090

Q1.

which system is better?

A. 2080 ti x 3way ( Vram 11gb x 3 =33gb)

B. 3090 x 1way (Vram 24gb )

Q2.
Does system A 's VRAM efficiency reach 33GB arround?

is it Larger than single 24gb??

My plan was 3080 ti x 3way or 4way (in March 2021)
But 3090 single has so problem ,now

by elon
Mon Jan 04, 2021 2:21 pm
Forum: Installation Support
Topic: [Guide] Using Faceswap on Nvidia RTX 30xx cards
Replies: 124
Views: 201047

Re: [Guide] Using Faceswap on Nvidia RTX 30xx cards

bub.JPG
bub.JPG (165.86 KiB) Viewed 42582 times

Somtimes it works fine . more than few hours with H128 model

but sometimes, freezing In few second, or in Few minute.

by elon
Mon Jan 04, 2021 8:05 am
Forum: Installation Support
Topic: [Guide] Using Faceswap on Nvidia RTX 30xx cards
Replies: 124
Views: 201047

Re: [Guide] Using Faceswap on Nvidia RTX 30xx cards

Ok figured it out under Trian = settings = orignal tick Allow growth. Seems to be processing will let you know the outcome. CUDA ERROR = I think it is VRAM problem. reduce batch size and input output size. Try DFL H128. and do not any other program even Chrome. and reduce batch size gain and agian ...
by elon
Sun Jan 03, 2021 1:53 pm
Forum: Installation Support
Topic: [Guide] Using Faceswap on Nvidia RTX 30xx cards
Replies: 124
Views: 201047

Re: [Guide] Using Faceswap on Nvidia RTX 30xx cards

ok managed to tick Allow growth which seam to work for extraction. During Training now getting: E tensorflow/stream_executor/cuda/cuda_dnn.cc:336] Could not create cudnn handle: CUDNN_STATUS_NOT_INITIALIZED 2021-01-02 19:56:38.723463: E tensorflow/stream_executor/cuda/cuda_dnn.cc:340] Error retriev...
by elon
Sun Jan 03, 2021 8:54 am
Forum: Installation Support
Topic: [Guide] Using Faceswap on Nvidia RTX 30xx cards
Replies: 124
Views: 201047

CUDA Error with rtx3090

2021-01-03 15:15:56.811830: E tensorflow/stream_executor/cuda/cuda_event.cc:29] Error polling for event status: failed to query event: CUDA_ERROR_LAUNCH_FAILED: unspecified launch failure 2021-01-03 15:15:56.811842: F tensorflow/stream_executor/cuda/cuda_dnn.cc:189] Check failed: status == CUDNN_STA...
by elon
Sat Jan 02, 2021 6:29 pm
Forum: Training Discussion
Topic: Too Many B Faces? 9000 pcs 512size
Replies: 4
Views: 2820

Re: Too Many B Faces? 9000 pcs 512size

Short answer: Too many similar faces does not help. You want variety. Q1 if I remove 6000 B faces from 9000 B faces, Can I save more Vram? Q2 and if i reduce 512 size ---> 256 size ,, Can i save more Vram? Q3 and I want Higher resolution converted result. but 128x128 is too low resolution how can I...
by elon
Sat Jan 02, 2021 4:31 pm
Forum: Training Discussion
Topic: Too Many B Faces? 9000 pcs 512size
Replies: 4
Views: 2820

Too Many B Faces? 9000 pcs 512size

I think,
I extracted Too many B faces (9000pcs+ 512size)

and most of faces are similar to each others

is it bad for training?

I am using rtx 3090

Removing more similar faces are helpful for training?