Comments

Blacklisted
    (help)
    Show Search Options
    post #14529
    269 435 2 E
    post #101812
    81 113 1 E
    post #101849
    135 249 1 E
    post #101733
    90 162 2 E

    OwOBuldge69

    Member

    I love how smooth this is x3! Any tips on maming these and what do you use? Been meaning to ask around for some help ^^"

  • |
  • 1
    •
    post #101724
    199 359 5 E

    OwOBuldge69

    Member

    I love this animation so much! <3 What do you use to make these animations and any tips on making them? ^w^"

  • |
  • 0
    •
    post #101784
    175 267 1 E
    post #92479
    138 233 5 E
    3 5 1 S

    ohnowhat

    Member

    Had a convo with Scrxtch a while back on Frosting about uploading a couple of their posts here
    They were actually the one who asked me, & seemed pretty chill about the possibility of the uploads getting deleted

    What a Legend

  • |
  • 0
    •
    post #101805
    248 407 2 E
    post #101814
    67 80 2 S
    2 10 1 E
    post #101817
    239 375 5 E
    post #101826
    33 45 2 S
    post #101826
    33 45 2 S
    post #101820
    94 151 2 E
    post #100955
    142 222 3 E
    post #101741
    85 166 4 E

    CartoonLover

    Member

    anonyfur89 said:
    Amazing!! Is this LORA available on CivitAI? If so, could you provide a link?

    Oh, I trained this lora recently, it's not avaliable anywhere.

  • |
  • 0
    •
    post #101694
    16 21 1 S
    post #101783
    107 159 2 E
    post #101761
    106 171 2 E
    post #101817
    239 375 5 E
    post #101817
    239 375 5 E
    post #101722
    200 321 4 E
    post #101754
    151 224 1 E
    post #101727
    96 131 1 E
    post #101351
    136 197 7 E
    post #101768
    343 536 8 E
    post #101741
    85 166 4 E
    106 154 2 E

    Lordgriffin

    Member

    I really do wish more people would generate images like this...even the Dark side needs to win. I really enjy this one

  • |
  • 0
    •
    post #101798
    150 289 4 E
    post #101681
    364 674 1 E
    post #101768
    343 536 8 E
    post #101775
    93 144 1 E
    post #101796
    109 215 1 E
    post #101805
    248 407 2 E
    post #101801
    105 187 2 E

    Nagiman90

    Member

    You have the right to be my bitch. Anything I do or Say will be followed. You have the right to lube as needed.

  • |
  • 2
    •
    post #101803
    157 273 2 E
    post #101803
    157 273 2 E
    post #37998
    304 482 3 E
    post #9080
    334 578 4 E
    post #101779
    179 290 7 E

    badwolf1

    Member

    Love the pussy on this one, a combination of dog and human.
    Looks anthro bottom, feral top.

  • |
  • 4
    •
    post #101745
    45 55 6 E

    Silvicultor

    Member

    b1techienne said:
    I have SageAttention, didn't know it would use it by default.

    However, despite the increased video length, it still seems to fail at the test of getting an actual progress in the video. Have you tried getting a character to undress themselves? For some reason, no I2V implementation of Hunyuan Video seems to be able to do this, I never got it once. It's like the model is able to add movement to what's present in the picture, but never to show anything new.

    Actually managed to install sageattention on Windows thanks to a helpful Github user. Gives a nice speed-up, but RAM usage is still absurdly high for me. But nvm, I have the RAM it can use it...😅
    No didn't try to make a character undress yet. Maybe that's one of it's limitations, because it always only generates one frame at a time. While it has some logic implemented to get context from the other frames, the context in longer videos is probably still it's weakness.
    It can however also be an advantage that it doesn't change so much. Other video models often introduce weird stuff and erratic movements that I never prompted for. This model doesn't. Didn't have a single output so far with "crazy morphing". The movements are often only small and slow - but correct.
    So even if you can't do everything with it, it is very interesting for the furry community because I would say it's the one of most uncensored video models we have.

  • |
  • 0
    •
    post #101765
    120 204 1 E
    post #101750
    127 222 1 E
    post #101745
    45 55 6 E

    b1techienne

    Member

    silvicultor said:
    Maybe, yes. I assume you installed sageattention that could help. As far as I know nobody on Windows managed to run FramePack with sageattention. For me it uses >50 GB of RAM, not a real problem tho, because I have 64 GB, but still that seems pretty crazy...

    I have SageAttention, didn't know it would use it by default.

    However, despite the increased video length, it still seems to fail at the test of getting an actual progress in the video. Have you tried getting a character to undress themselves? For some reason, no I2V implementation of Hunyuan Video seems to be able to do this, I never got it once. It's like the model is able to add movement to what's present in the picture, but never to show anything new.

  • |
  • 0
    •
    post #101286
    322 514 7 E
    post #101745
    45 55 6 E

    Silvicultor

    Member

    b1techienne said:
    Weirdly, it doesn't seems to take much memory here, it's like ~16gb... I'm using Linux, maybe they do things differently?

    + https://e6ai.net/posts/101780

    It's incredibly fast too. That's an unexpected surprise

    Maybe, yes. I assume you installed sageattention that could help. As far as I know nobody on Windows managed to run FramePack with sageattention. For me it uses >50 GB of RAM, not a real problem tho, because I have 64 GB, but still that seems pretty crazy...

  • |
  • 0
    •
    post #101768
    343 536 8 E

    Hogan

    Member

    weeml said:
    What a lovely mix of realism/detail and style

    Seconded! This is amazing, would looooove to see more of her!

  • |
  • 3
    •
    post #101745
    45 55 6 E

    b1techienne

    Member

    silvicultor said:
    Yes, it is really an interesting project. But one important info is missing in the readme of the repo: Yes, you can run on very low VRAM because it only generates one frame at a time, but then you need big system RAM to offload. The used model is >30 GB!

    Weirdly, it doesn't seems to take much memory here, it's like ~16gb... I'm using Linux, maybe they do things differently?

    + https://e6ai.net/posts/101780

    It's incredibly fast too. That's an unexpected surprise

  • |
  • 0
    •
    post #101701
    135 221 3 E
    post #101780
    59 82 2 E

    b1techienne

    Member

    lemlem said:
    mf be sippin no care in the world

    No care, no world, only sucking the precious boob juice and feeling the motherly warmth of this generous gazelle lady.

  • |
  • 6
    •
    post #79602
    73 103 5 E
    post #101780
    59 82 2 E
    post #101745
    45 55 6 E

    Silvicultor

    Member

    b1techienne said:
    This sounds really interesting. And the results seems promising, too

    Yes, it is really an interesting project. But one important info is missing in the readme of the repo: Yes, you can run on very low VRAM because it only generates one frame at a time, but then you need big system RAM to offload. The used model is >30 GB!

  • |
  • 0
    •
    post #101779
    179 290 7 E
    post #101768
    343 536 8 E
    post #101779
    179 290 7 E
    post #100820
    154 258 2 E

    Knotsfordays

    Member

    OMG this is the hottest thing ever... Is there any chance you'd be willing to share a prompt for this?

  • |
  • 3
    •
    post #101763
    215 323 1 E
    post #101600
    298 562 2 E
    87 128 3 E
    post #101768
    343 536 8 E
    post #101774
    128 243 1 E
    post #101600
    298 562 2 E
    post #101745
    45 55 6 E

    b1techienne

    Member

    To generate 1-minute video (60 seconds) at 30fps (1800 frames) using 13B model, the minimal required GPU memory is 6GB. (Yes 6 GB, not a typo. Laptop GPUs are okay.)

    This sounds really interesting. And the results seems promising, too

  • |
  • 1
    •
    post #101768
    343 536 8 E
    post #101307
    203 325 5 E

    BirdLover

    Member

    mikubraver said:
    Can I carry her in a barrel full of alcohol, like on Baldur's Gate 3. xD

    Of course. But prepare to fight when you release her. But not in the usual way :> :> :>

  • |
  • 3
    •
    post #101761
    106 171 2 E
    post #97501
    281 398 15 S

    Raikou239

    Member

    cuyi said:
    They miss Renamon IMO

    This is definitely biased toward modern, cuz I too would include Renamon and Lola.

  • |
  • 0
    •
    post #101768
    343 536 8 E
    87 128 3 E
    post #101418
    231 372 1 E
    post #101715
    198 293 2 E
    post #101698
    24 19 3 S

    Alpha~Centaur

    Member

    This looks like a real video game.
    I wonder if programmers could use AI to generate all pixel art in a game...
    (probably not for animations)

  • |
  • 1
    •
    post #101621
    175 274 4 E