Thread 5865556

Previous thread >>5835060
Dedicated Suno/Udio/Sonauto thread >>5864318

Post AI generated stuff. Song covers, animations, whatever.
Bonus for OC. Please keep 6 second Hunyuan/Pixverse/Kling clips without sound to a minimum.
Open source resources are linked below, but feel free to post anything, open or proprietary.

> Speech-to-Speech and Singing-Voice-Conversion
https://github.com/Mangio621/Mangio-RVC-Fork
https://github.com/Vali-98/XTTS-RVC-UI
https://github.com/voicepaw/so-vits-svc-fork
https://github.com/open-mmlab/Amphion/tree/main/models/svc/vevosing
https://github.com/Plachtaa/seed-vc

> Text-to-Speech
https://github.com/index-tts/index-tts
https://github.com/Zyphra/Zonos/
https://github.com/open-mmlab/Amphion/tree/main/models/tts/maskgct
https://github.com/SparkAudio/Spark-TTS
https://huggingface.co/spaces/mrfakename/E2-F5-TTS/tree/main
https://github.com/collabora/WhisperSpeech
https://github.com/myshell-ai/OpenVoice
https://github.com/yl4579/StyleTTS2
https://github.com/BoltzmannEntropy/xtts2-ui

> Text to Music
https://github.com/joeljuvel/YuE-UI/

> Text-to-Video, Image-to-Video
https://github.com/kijai/ComfyUI-WanVideoWrapper
https://github.com/kijai/ComfyUI-CogVideoXWrapper
https://github.com/kijai/ComfyUI-PyramidFlowWrapper
https://github.com/kijai/ComfyUI-HunyuanVideoWrapper

> Deepfake and Lipsync
https://github.com/Hillobar/Rope
https://github.com/Mozer/wav2lip
https://github.com/saifhassan/Wav2Lip-HD

> Audio Cleanup
UVR Walkthrough: https://docs.google.com/document/d/17fjNvJzj8ZGSer7c7OFe_CNfUKbAxEh_OBv94ZdRG5c/edit#heading=h.n8ac32fhltgg
https://github.com/Anjok07/ultimatevocalremovergui
https://github.com/Rikorose/DeepFilterNet
https://github.com/resemble-ai/resemble-enhance

> Transcription
https://github.com/linto-ai/whisper-timestamped
https://github.com/openai/whisper

May the 4th be with you. What Star Wars stuff do you got?

>>5865690
Better than the sequels.

>>5865690
What happened to that anon that posted Darth Vader stuff every once in a while?

Read Siege by James Mason

>>5866678
>>5866673
These are great. Have any more?

>>5866738
there are loads
https://www.instagram.com/timeless__snaps/

>>5867722
The cuts are really obvious, AI should incorporate video-to-video momentum consistency.

>>5868612
the records are like frying pans, because the music is so hot
and the piano is like a typewriter
haha so funny

>>5866843
Being Italian these fucking things are even worse, it hurts inside.

post that high quality AI slop if u guys have any, much appreciated.

>>5870256
sloppa sloppa

>>5871441
I hate this

>>5871441
It certainly captured the feel of the show. TV is still abhorrently bad though.

>>5872172
>[Konosuba AI] Megumin English Dub but it's Rie Takahashi
>September 23, 2023
>This RVC Over-Dub shows a scenario that I have been curious about, what IF It's the anime Dub but with the characteristics of the JP Voice Actor/Actress, and it turned out quite nice! Although because Rie is Japanese there is some of that "engrish" going on that makes some of the sentences harder to understand. REMEMBER THIS IS JUST FOR ENTERTAINMENT!

>Meguminv2 Model by: Thern
>Tool Used: RVC V2 and UVR v5

Anyone got that about warwick davis letter to 4chan about him getting bullied?

Haram

>>5873847
The Monolith lies

>>5876861
Tiggy love

>>5873847
>>5876861
>>5877532
WTF loving those

>>5877532
Justice for Snarbo

>>5877610
In that case, have another

>>5878264
WEW
Mildly funny, more scary than funny. Unsettling.
Like a fever dream I know shouldn't be real, but have to accept is, because I can't wake up.
The remaining 131 should fit ITT, la, keep em coming.

>>5878313
I would unironically watch a trailer park remake of GoT

Can anyone tell what this is.. AI, CGI, live action..

>>5878489
blender, probably

>>5878489
Fake AF, since no lorry would stop this fast.

>>5865560
KEK

Chat is this real?

>>5868612
I guess this was made from a pre-existing cartoon, but it's pretty well done and some kind of technical achievement for amateur gen (I suspect the big studios already have the tech available for themsleve s to make pretty much anything they want)

>>5871441
pretty good

>>5878489
My guess is on some well done stop motion and movie editing with real live action cameras

>>5880353
I have no trouble believing he'd say this, and he'd gain massive support for saying what we're all thinking.

>>5880350
damn that's some really low-grade slop.
have something actually cutting-edge as a palate cleanser

>>5880605
no way, this is 100% AI?

>>5880617
yes, not even edited apart from the separate scenes being combined (5 mins in video editor)

>>5880621
oh and the text and logo at the bottom are probably added, but again 5 minute job

source:
>I used to shoot $500k pharmaceutical commercials.
>I made this for $500 in Veo 3 credits in less than a day.
https://x.com/PJaccetturo/status/1925464847900352590

>>5880621
>>5880624
Fucking crazy, this is the first impressive AI vid I've ever seen.

>>5880605
They need to improve the audio, it sounds like it's coming from a sound booth rather than an echoey room or an open space. It doesn't blend with the environment. Also it's not 100% timed with their lips

>>5881020
sorry, this one is a lot more compressed looking than i'd like. 6MB limit for 2:25 is pretty rough though

>>5871441
SOVL

>>5873847
VEO2?

>>5880350
undertale

>>5878264
>>5876861
dude these are so good lol

Anyone got some VEO3 stuff?

>>5881020
This is great wtf

'member dignifai? i 'member

>>5881085
see
>>5879403

>>5881776
I rather like this. It shows how good all these degens could look if they weren't shitting themselves up.

>>5881793
>>5881791
no way those aren't bots

>>5881759
Jfc it won't be long now... Before nothing online is real.

>>5881759
We aren't "ones and zeroes", but there are (considering current knowledge) completely valid interpretations of reality in which we are ultimately equivalent to them.
The ultimate question is, if that's the "ground truth", but it is certainly a reasonable theory to consider.

>>5881905
A pet theory of mine is, that reality isn't "real".
Reality is probably only a countable infinity - unlike the *real* numbers that are an uncountable infinity.

>>5881910
Even if we're all in a "simulated" universe it doesn't have much practical meaning from our perspective. There is no way for you to get "outside" even in principle.
It's roughly analogous to the inability for "you" to escape your conscious experience. Even the influence of information from outside is necessarily actually experienced as a representation coming from inside.

>>5881728
>>5881731
They finally did it. They finally sucked the last remaining soul out of AI video. From now on it's soulless and in the uncanny valley.

>>5881921
"Reality" exists independent of a concrete calculation - exactly like all the digits of pi exist, if you are actively calculating them or not.
That doesn't mean, "reality" is not ultimately equivalent to a calculation, though.
I think, Wolfram's idea of a hypergraph is very interesting in this context: It neatly explains, why we don't perceive the universe as discrete.
Automata theory in general is a big eye-opener. I'd recommend everyone to have a look at it at some point in their lives.

>>5881926
That webm was the golden age of GAI.
I'd rather like to see produce like that instead of what you answered to.

I hate the future so fucking much

>>5882084
its beautiful and you WILL use your new mandatory AI assistant.

>>5881029
>>5881735
https://www.youtube.com/watch?v=unfpnIF0OMo

>>5881910
>I think, therefore I am.
Does it really matter if we are simulated? We have the ability to think as individuals and therefore have an ego, a sense of true existance even if the "reality" we live in is simulated.

>>5882114
The crucial observation precisely is, that it does not matter.

I'd love to see the day where we successfully prove that reality isn't "real" (as in the real numbers).
My argument is, that since you can store infinite information in a single real number, the result of reality being "real" would be that everything becomes black holes.
We don't observe that, so reality isn't "real".
Unfortunately this way of thinking appears half-assed and not thorough enough. It also seems way too simple for actual physicists not having thought about and dismissed it.
I have no proper high-level education in physics, so I'm most certainly missing something.

>>5882120
This nigga all deep but he don't know bout my noumena n shieet

Titanic

>>5881776
Man.. that's awesome. Also a huge trolling potential. Imagine how the world could be if we harrassed thots with "Imagine what you could have been instead"-images

>>5883012
That happened at the time and as you can imagine, hoes mad.

>>5880353
Trump finally taking on the most evil conspiracy

>>5883084
most mentally sane tranny generation, lol

2 years of difference

>>5883695
I gotta get into WAN. I'm just intimidated by the RAM/VRAM usage.

>>5883725
if you dont have a powerful gpu with 20+GB, just stock up standard ram. A used DDR4 64GB is only ~$50 on ebay.

>>5883015
Brilliant

>>5883732
I have 16GB VRAM and 64GB RAM, but don't most workflows need a combined 80GB? My job's about to scrap some computers in a couple months and I might be able to get my hands on 1 maybe 2 20GB Nvidia cards so maybe I'll be able to get into it later this year.

>>5883744
Just double the ram. Its the cheapest option. Alternatively, I think if you got pagefile/swap that works too. But thats slow.

>>5883747
Yeah you're probably right. It's just a damn shame that my sweet expensive low latency RAM isn't enough...

>>5866678
Moar Egyptian girls all day 24/7/365.25

>>5871441
>The one with the tigers
My fucking sides

>>5878264
>>5876861
>>5873847

i like these

>>5883744
I just did a quick vram usage test with WAN VACE 14B Q6. I was using ~60GB of ram. I think you should double the ram to 128GB for comfortable margin imo.

>>5884239
>128gigs
Not him, but isn't that several highbend GPUs worth of ram? I remember a guide saying you need an ass ton of VRam for good rendering.

>>5884361
Oh I meant, just use standard RAM. 128GB is cheap as fuck as proofs you against video creation and its long ram usage.

>>5884361
It can do CPU offloading the same way that LLMs do so you just need to total up the RAM+VRAM. Of course VRAM is faster so the more of that the better.
>>5884239
Ok thanks, maybe it's not hopeless with my current setup.

>>5881776
DignifAI is the first install on my AR glasses if they become powerful enough.
2nd install is Makeapp

>>5879403
People can talk, got it, but why add blacks? What's the joke here? Implying they are people too?

>>5886386
Did you reverse the video halfway through? The movement of the hair in the second half looks off.

>>5887002
Google AI. It will modify your prompt to explicitly include a quota of black people and you won't be able to turn it off.

>>5885498
this is fucking great lmao

>>5881759
>>5881905
>>5881910
>>5881910
the joke is that that video is AI, it never happened those people

>>5865560
>only the pink one has a face in the end for some reason
fucking lol

>>5868612
man I can't believe I actually remember every single detail from that episode of tom and jerry. but I still laughed out loud because of the rudimentary mistakes AI still makes on those animations. It's still fucking impressive obviously

>>5871441
this >>5883877 holy shit
people hysterical about two tigers eating a married couple alive

shit but this whole thing was awesome

>>5873847
holy shit I can't remember the last time I laughed so hard

>>5878264
>>5877612
>>5877532
>>5876861
>>5873847
>>5866678
>>5866673
>>5866843
>>5865592
>>5865560
All fucking awesome, still have half a thread to lurk.

>>5880519
How about this one?

>>5881776
Remember the one that stripped bitches of makeup?

>>5886386
nah
I'd win

>>5887907
Ha, very nice!

>>5880353
Poop is poop folks

>>5881776
antesedepois federal edition when
is there a if_you_weren't_fat app?

>>5882084
To be fair, it's not the worst way the world could have ended.

>>5887757
>implicating anyone you are quoting wasn't aware of that

>>5885498
Holy shit this is great.

>>5887907
got me laughing in the first 2 seconds lol

https://www.youtube.com/watch?v=WJhrgiVf51k

Could I interest you all in some Peter Pubes?

very happy with how this came out

>>5889619
Don't mix in that Hebrew nonsense

>>5889723
So angry.

>>5889389
You mean this was made with AI?

>>5868612
If you pause it at 41s right as the pancake carrying cat is on the left side of the orange one, it has a human face.

>>5890208
This is what bothers me most.. we're in a post-truth world.

>>5890217
>truth
That term is worthless.
They are trying to suggest their "truth" and reality are the same thing.
You know, the people that try to tell you that males can be female and that if you disagree, you are a Nazi.
Their "truths" for the most part aren't real. Which doesn't mean their enemies are the good guys by default, but it can't be too wrong to side with reality.

>>5890217
>muh truth
An age of storytelling. Here's a big fabrication: we both have a mind. Even though it's not real dude, you're making yourself up. IMAGINATION

>>5890169
Yes

>>5890268
Then it's very amazing.

To which extent did you have to give directions of which chord or general sound to be like at each given part?

>>5890268
>>5890476
(I just want to add:
I don't even appreciate this kind of music, and I've heard it on repeat at least 3 times - I had no idea AI already allowed us to create music like this)

>>5890476
Just suno and a bit of lyric editing.

>>5890478
... have you seen the thread full of AI music?
>>5864318

>>5866673
>bro took it personally :(
because he died. fucking loled

>>5890485
I surely will.

I love following all about AI and I have absolutely dreamt about producing music in a more sophisticated software that allowed me to interact with it without having to master music theory, for years.

I can't believe I'm only now delving into the knowledge that we can produce music so easily right now. Thank you, will check it out tomorrow for sure.

Just a question by the way, this shit is all still expensive, right? Or can anyone build specific intrumentality for free as it is?

>>5890535
Most solutions are proprietary with a freemium model. The free and open source stuff is still under construction.

>>5887907
let him sing the whole Nigga Heil Hitler song from Kanye West

>>5890238
I'd pay to watch this in the theatre, legit, no cap bro

any cultivation enjoyers?

>>5865560
holy cinema

>>5865560
the random inconsistencies is what gets me
>red has a white glove independent from his suit
>yellow, green, and pink have no glove
>blue has "gloves" connected to his suit

>>5890238
>Pass me that dinglehopper, I've got some new hair to comb
Unbelievably kino

generating rare merchants with AI

a TV documentary from the future where Toki Pona is the common tongue

>>5887907
Is this a genuine request? If so...
>President Donald J. Trump here, just letting you know that I just signed an executive order putting YOU into the Baby Book. Forever.
>And in addition to that, I am taking anyone you previously put into the Baby Book and taking them permanently out of it and granting a forever immunity to those people from ever being put in the Baby Book ever again.
>If you dare challenge this order you will be deported to the Land of the Poop. Forever.

>>5878489
those clouds move as if it's a badly done video edit, but the rest is breddy good video edit stuff, not AI sadly

>>5878489
Its from old info campaign spreading awareness about speeding on roads. man made production funded by state.

>>5880107
I think it's Top Cat.
An old cartoon.

GOOD MORNING SAARS

>>5867182
how is the peephole moving like that?

>>5892527
If the vignette was moving I'd say it's an illusion caused by the wide angle lens

nice

>>5872002
what model?

>>5893125
Vevo, made by the same people who made MaskGCT. It only needs 15-30 seconds of reference voice. Neat but has its drawbacks.
https://github.com/open-mmlab/Amphion/tree/main/models/vc/vevo

I made a version that integrates Ultimate Vocal Remover and auto segments anything over the max allowed length.
https://github.com/chameleon-ai/redubber

>>5893221
cool stuff man, i've only messed with i2v/v2v with default or music audio. so adding/changing the voice dialogue instead to go with it looks fun.

>>5894651
Holy Kino

>>5895038
this is gud

MORE PLEASE

>>5895038
>The Internet won't last more than a year.
The Internet was already more than 20 years old at that point.

I'll kinda miss the schizo, dream like AI now that everything is becoming more and more real

>>5895817
Remember the super early generative AI that could only produce trippy shit with seal faces hidden throughout? I miss that stuff. Hard to even find it now

>>5895929
Oh man I'm trying to remember the name of that one, it was called "dream machine" or something. It would do psychedelic patterns and shit.

>>5896061
deep dream! Thanks anon. Crazy how hard it is to find now that modern generative AI has emerged.

>>5895034
>>5895038
first impression was "this looks great, but it's wasted on shit writers"
then i realized, is it written by ai too?

>>5895929
Remember what they took from you

>>5896196
Might as well, especially for parody where the aim is to imitate an existing format. That's ideal use case for AI

There is some funny shit in this
goddamn

>>5896061
>>5895929
>>5896178
even crazier is that it was only like 5 years ago

>>5896346
The god damn dogs everywhere

Just finished re-making the Fist of the North Star music video with Phase Girls. Too big for my brain to webm though

>>5896178
I found it! I'm gonna try to run it.
https://github.com/google/deepdream

>>5896559
catbox it and I'll convert it or try
https://github.com/chameleon-ai/webm-for-4chan

I need that one ai voice slop video of Saruman reciting the Rod and Ring pasta from 2016, anons.

Anyone have the series where the black guy leaves his white wife who just gave birth, steals a bike, and goes to KFC?

Whole Fist of the Phase Star trailer reshoot with phase girls.

Learning Wan has been fun. Video editing not so much.

I shrunk her as low as she will go

>>5896681
Nice. How did you do it, like did you start with Stable Diffusion gens and ran it through img2video or what?

>>5896685
That's the sum of it. I will admit i suck at art both ai and not though. For images i used Illustrious. At first i used control nets but i switched to a multi mask workflow a coomer showed me. I went from 30 mins of tinkering with poses to 5 min an image, heavily recommend looking into multiple mask workflows.

I2V was Wan. Its very straightforward once you nail down the settings and learn the way it likes prompting. I would say 40% of my gens are usable which im very happy with because stock i feel like its 5% at a very slow speed.. There are also lora's for specific camera shots it can do which helped. It does not really like anime but it "can" work if your ok with a 3d style.

I think i could do this video over again in a day and make something better now that i understand the systems. Im looking into a longer video to make next, maybe Phase Saw man but im unsure.

Paid stuff blows it out of the water without heating your house up though.

>>5872002
PANKO NO

>>5896742
I'm gonna have to mess with it because I was waiting for a usable i2v for like a year and it seems like it's finally time, or at least Wan is on the right tragectory.
Last thing I used successfully was AnimateDiff V3 and tried cogvideo with much failure before Wan was a thing. I'm decently good at SD so my missing piece was i2v.

>>5896606
I got this example working on python3.10 (tried newer pythons and they're a no-go)
https://www.tensorflow.org/tutorials/generative/deepdream
but all the old DeepDream repos are dead (they're 10 years old!) and use python 2. I wonder if I can replicate the video technique. That would be neat to resurrect it for old time's sake.

>>5896880
There's no way to force temporal coherence in those old algorithms. You would just have to apply it to each frame individually, at least im pretty sure that's what they did for >>5896346

>>5896894
Although I'm just assuming that since the result is so noisy

>>5896880
Here's a neat example from their gallery, linked on the github page

>>5896894
Yeah I can just split the video into frames, but this guy says that he feeds the output into the next input for some semblence of coherence, kind of like old Stable Diffusion Deforum would do.
https://github.com/graphific/DeepDreamVideo/wiki

>>5896904
Ohhh. That makes sense. Pretty straightforward too

wip track

>>5896742
regarding paid stuff, do you have a recommendation for paid model that can handle anime really well?

>>5896944
Google Veo 3 is probably the best i have seen. I have not looked much into it though(dont want to spoil myself on the quality difference)

>>5896772
If you like this level of animation i recommend it. Its not great but its fun. It does real people way better than anime. With my 24gb card I made the attached clip in 180 seconds of generation time and optionally passed it through a different workflow and did a 120 second x2 upscale on it(I gen in 640x480 for speed). I can share workflows if needed for comfy.

>>5891386
Pretty cool, anon. I think "misinterpreting the world" would work, too.

>>5865560
Surprised no one else mentioned that this is funny because japs are running out of rice with prices going crazy that they have to dishonorably import us, chinese, and korean rice.

>>5896981
thanks fren, will take a look at veo3

>>5896981
Yeah man if you'd share your comfy workflows that would give me a head start. Do you think I'll be alright with 16GB VRAM + 64GB RAM or am I gonna OOM?

>>5896880
Got naive rendering done where it's doing individual frames without blending. Now I have to learn how to feed the output image forward.

>>5897377
I would guess you could just mix the generated current frame (f1') with the raw next frame (f2). Like a*f1'+(1-a)*f2. Not sure what percentage though. Like a=1-10% I guess.
You could try separating out the AI generated component of f1 before adding, so like a*(f1'-f1)+b*f2.

>>5897450
typed b instead of 1-a
oops

>>5897450
I'm going to have a configurable amount, I know that much. As for the actual blending, the images when loaded are like...numpy matrices or something. I don't really know how to translate that into image manipulation so I gotta educate myself.
Separating out the AI noise is interesting. I didn't consider it before.

>>5897460
Numpy arrays are super easy to work with. You ought to be able to simply add and subtract them just like numbers, as well as multiply them by constants. For an image, the dimensions are like [width, height, RGB] or something, so [1000,2000,0] would be 1000th row, 2000th column, red component. Maybe I have width and height backwards, and RGB might actually be ordered like BGR. I cant exactly remember how numpy handles images, but it shouldn't really matter if you use numpy to convert from file to matrix and also from matrix back to file again

>>5897657
Shit. Adding and subtracting them might give you negative numbers or numbers above the max pixel value. I suppose you could just clamp them.

>>5897662
I didn't worry about it too much after I looked at what the DeepDreamVideo dude did. Before converting to numpy arrays, they're pillow images, and pillow has "blend" and "difference" functions. DeepDreamVideo used blend but not difference. I quite like the artifact persistance you can get from blending the difference. That was a great idea. This is with a 20% blend.

https://github.com/chameleon-ai/deepdream-revived

>>5897675
That looks pretty good. Seems to work well, but I see there's a static overlay pattern for some reason.

>>5897775
Actually, i see that static pattern is only on parts of the scene that are solid black. I guess it makes sense that the pattern won't change if there's no information in the image

>>5896904
I loved these... Shame that technology is already in the past

>>5897779
It also seems that all the "eyes" are the same size, so maybe somthing to tweak?

>>5897775
It could have something to do with how it's processing the image in 512x512 tiles. That's stuff I just copied from the tutorial without looking at too hard. With some studying I can probably improve the tiling behavior.

>>5898117
I'm gonna have to get into the deepdream game because I just realized there's probably a lot of possibilities that never got explored because it's relevance passed so quickly.
I woke up this morning thinking about how you could add noise to the image to influence the deep dream result and then subtract it after processing.

>>5898325
You could probably apply a lot of modern techniques used by Stable Diffusion. I was reading up on it and it's basically a very early version of diffusion based image generation, just with no way to prompt it.
https://research.google/blog/inceptionism-going-deeper-into-neural-networks/

What is considered the best text-to-speech for Windows?

>>5898706
Microsoft Sam

>>5898713
You have selected Microsoft Sam as the computer's default voice

>>5898874
>my roflcopter goes soi soi soi soi soi soi soi soi soi soi soi soi soi soi soi soi soi soi

>>5896061
I can save her, with my penis, but why Zelda dungeon music?

>>5898893
https://www.youtube.com/watch?v=2J6wAiHQ5U0

>>5898706
Entirely depends on what you want to do with it.
I've used espeak in the past and it did its job. Offline. Without any cloud-bullshit.
That said, I have a soft spot for Sam: >>5898713
It felt amazing to use that fucker in the 90s.

>>5890217
I solved it by focusing on context in which something is said or posted, and more importantly degree. Many times something is said, but the degree does not match. I stay comfy, if slightly more ignorant.

>>5900020
lmao

>>5890217
Is that true?

>>5873018
PAY VOICE ACTORS

extended the image in chatgpt and made 2 additional frames

>>5900836
>pay for something that's free now
LOL

>>5871441
Incredible

>>5900976
>and made 2 additional frames

>>5900786
Fiddle coon and banjo squirrel stole the show

>>5889723
suddenly, anon found himself the boss of me

>>5890217
What truth? The one written by the victors in the history books?

>>5881759
Can someone redo this but replace "prompts" with "scripts"?

>>5901008
parasite retard

>>5904228
Seethe harder fag

>>5878057
this is an AI thread, not real footage of jews thread

>>5883084
the faggots who make and spam this shit were absolutely shocked when kamala lost

>>5905195
kek

>>5895034
>>5895038
cool, but also cringe

>>5893251
kek

>>5903620
lmaooo more please

>hyper-realistic """AI"""
Someone figured out how to plug a shitty LLM AI to Unreal Engine or similar, eh?

>>5873847
>>5876861
>>5877532
this one is probably my favorite joke of the entire series

>>5906087
forgot my video

>>5905829
You're never going to get laid with that anti transhumanist attitude.
>None of you guys ever get laid anyway!

With all this palantir talk lately and now war with iran, this sounds like a good moment to make a video version of this: https://vocaroo.com/19jLJHZHYJTw

>>5905829
This goes hard

>>5903615
kek

>>5867722
>hooked up to a neuralink playing AI generated horror 24hrs in a CIA blacksite with feeding tube

>>5906355
>country people are so unhappy
lmao incredible projection

>>5905195
Good slop

>>5906532
More/source? I can only find a few on youtube and there doesn't seem to be an original uploader

>>5906565
gossip.goblin on tiktok

>>5906627
>no diamews in the sky song

>>5906627
based Garfield

>>5906532
smoke weed lmao is kind of a step down after this one >>5905829

>>5906898
who asked? i don't recall anyone asking for your advice.

>>5906924
Me. I asked.

>>5906924
imagine being this butthurt over a tiktok video

>Voice2Voice RVC
When I add a model that has both the model.pth & model.index it barely progresses and uses cpu. With just the model.path it uses gpu and finished quickly. Why is this? Aren't I supposed to use both?

>>5906956
yeah I know, they're pussy for crying about it

>>5903620
i asked her why she was sucking down kike jizz with her now ex, and she blocked me :****(

>>5905829
>>5906532

shalom

>>5907559
Where is the full version of this one?

>>5908520
nvm found it here
https://m.youtube.com/shorts/MVR1lWW8rFY

Can I see someone take a real video and convert it into anime/cartoon style? preferably something close to the image at the beginning of this video https://www.youtube.com/watch?v=mGdXDL2300k