DF Community model tests: test #1 Batch size, all 5 clips comparison/compilation.
Duration: 7:08
Views: 9 321
Submitted: 1 year ago
Submitted by:
Celebrities:
Jennifer Connelly
Original Pornstar: Solazola
Description:
Here is a compilation of all 5 videos showing different batch size trained models along with a zoom in at 0:05 and 7 different static shots/frames from the video cycling between each of the 5 videos from 05:35 till the end. There was is an issue with rendering of the end sequence (static images comparisons) so please watch this video for a fixed version:
//deep.whitecatchel.ru/literotica/video/11344/df-community-model-tests-test-1-batch-size-end-sequence-fixed
Once you read this and watch all videos please vote for which one you think looks best here:
/>
Here is what we did, we've used the same model, src and dst datasets, we've applied the same workflow to how we trained those models, only difference between them is batch size of the model. To ensure each model had equal chance to learn properly we didn't train each model for the same amount of time but rather trained it based on epoch targets.
One epoch in simple words is done when a model processes all of the faces from a training dataset, in this case an average of SRC and DST dataset image count, since each model used different batch size it would process different amount of faces in a given amount of iterations.
- model trained at batch size 6 would do 6000 faces in 1000 iterations
- and at batch size 10 it would do 10.000 faces in 1000 iterations.
We've trained highest batch size model on each stage until we saw loss values stop improving and then we've calculated epoch/iteration targets for other models.
Here are full versions of all clips so you can watch them in full quality and resolution too.
Clip 1 and 3 by me:
//deep.whitecatchel.ru/literotica/video/11202/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly
//deep.whitecatchel.ru/literotica/video/11328/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-3
Clip 2 by //deep.whitecatchel.ru/literotica/user/145482
//deep.whitecatchel.ru/literotica/video/11331/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-2
Clip 4 by //deep.whitecatchel.ru/literotica/user/82429
//deep.whitecatchel.ru/literotica/video/11337/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-4
and Clip 5 by //deep.whitecatchel.ru/literotica/user/38333
//deep.whitecatchel.ru/literotica/video/11335/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-5
Here you can vote for which one you think came out looking best:
/>
//deep.whitecatchel.ru/literotica/video/11344/df-community-model-tests-test-1-batch-size-end-sequence-fixed
Once you read this and watch all videos please vote for which one you think looks best here:
/>
Here is what we did, we've used the same model, src and dst datasets, we've applied the same workflow to how we trained those models, only difference between them is batch size of the model. To ensure each model had equal chance to learn properly we didn't train each model for the same amount of time but rather trained it based on epoch targets.
One epoch in simple words is done when a model processes all of the faces from a training dataset, in this case an average of SRC and DST dataset image count, since each model used different batch size it would process different amount of faces in a given amount of iterations.
- model trained at batch size 6 would do 6000 faces in 1000 iterations
- and at batch size 10 it would do 10.000 faces in 1000 iterations.
We've trained highest batch size model on each stage until we saw loss values stop improving and then we've calculated epoch/iteration targets for other models.
Here are full versions of all clips so you can watch them in full quality and resolution too.
Clip 1 and 3 by me:
//deep.whitecatchel.ru/literotica/video/11202/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly
//deep.whitecatchel.ru/literotica/video/11328/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-3
Clip 2 by //deep.whitecatchel.ru/literotica/user/145482
//deep.whitecatchel.ru/literotica/video/11331/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-2
Clip 4 by //deep.whitecatchel.ru/literotica/user/82429
//deep.whitecatchel.ru/literotica/video/11337/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-4
and Clip 5 by //deep.whitecatchel.ru/literotica/user/38333
//deep.whitecatchel.ru/literotica/video/11335/df-community-model-tests-test-1-batch-size-featuring-jennifer-connelly-5
Here you can vote for which one you think came out looking best:
/>
Categories:
Straight Sex
Download now to support TMBDF
DF Community model tests: test #1 Batch size, all 5 clips comparison/compilation.
You currently have 0 tokens available
Insufficient tokens - 200 Required
Buy more tokensDonate to support TMBDF
Confirm tokens to TMBDF
1
2
3
4
5
6
7
8
9
10
11
12
![Anupama having sex in hospital [Teaser]](http://deep.whitecatchel.ru/literotica/contents/videos_screenshots/15000/15244/preview.jpg)
«Forum thread for further discussion: //deep.whitecatchel.ru/literotica/forums/thread-df-model-tests»
Reply
«There was an issue in the end sequence with static images, for some reason number on some clips where missing and also the actual images were not all static, here is a fixed version: //deep.whitecatchel.ru/literotica/video/11344/df-community-model-tests-test-1-batch-size-end-sequence-fixed»
Reply
«i was going to suggest a side by side. They all have an insane lazy eye which is very distracting but overall they all look awesome. Honestly cant tell which is best :?»
Reply
Reply to: Emmastone
1 year ago«Yeah... Actual side by side was not really possible, or rather not that easy (I'm lazy) so I've just did this. At the end, especially in the 2nd video I've uploaded after this is a fixed version of the end sequence that has full sized freezed frames so it's easier to judge each one, the number means the video/model/batch size but we're not revealing them until you guys (and we ourselves) rate them.
»
I know about the eye, bad src set but it was too late to change it when we've discovered it and we didn't want to add older pics of Connelly so we were stuck with her younger 90s version for which there aren't many good sources (only 2-3 movies).
Also eye priority was not planned in the workflow, we ended up running it but it didn't help much.
You need to choose one and vote, you can't give up that easily
Ответить