@@sanakchandraI bet you can get the same quality of photos on a flagship smartphone nowadays. Even studio quality photos might cost you max 10$. Where a professional photoshoot will cost you 500$ that's when the AI comes into the play.
@@sanakchandra, keenly see, in 24hrs we all look a bit different.. in whole week as well.. so, consider the photo to be clicked on your best day while on exam you are not so bright.. then also it's same difference.. just saying, it can happen.. in camera the tech is hidden, so we assume it to be called as natural.. while, in AI, we're addressing it as tech tech tech.. even camera is a tech, i agree it's literally light mapping on it's memory/paper.. but AI is similar, kind of a artist making live portrait of you, a bit off
Guys if any of u with coding experience can simplyfy this kind of thing to a point where this can be done with a friendly UI. Trust me u r going to make millions in the next AI bubble.
The output images will depend on the input image quality and the number of angles from which the image of the subject is taken. This is a very time consuming process. Or one can buy a very expensive CUDA capable graphic card and do the computation over there.
For those who don't know who Abhishek Thakur is, hes a champ on kaggle. Its like Michael Phelps but on kaggle and kaggle is jist olympics of Data Science and ML.
First Google colab started banning accounts using free service for AI related works , specifically Stable diffusion and SDXL model is around 6.5 gb and if you feel comfortable in downloading and uploading a 6.5 gb file to Gdrive of different google accounts then its okay to follow the process as Google will ban colab services for your account as soon as you get in their system's radar... and this is an year old recipe which is outdated now
true identity is an interesting word, we know the face.. not the spirit n soul, so we really lack knowing who's who.. so just another level of distortion by AI
when I run this it shows ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. tensorflow-metadata 1.14.0 requires protobuf=3.20.3, but you have protobuf 4.23.4 which is incompatible
Varun i training model on my images and save model file from generated folder. Can you please tell me what was the next step or what i do with this file?
@VarunMayya love your work but for the love of god tell your editor the sub-titles are competing with Ranveer Sigh - they are so everywhere so hard to keep track,
I'd go to the studio and get myself clicked instead of putting 2 hrs. Then run some errands and come back home with the edited photos already in my email.
Bruh, after struggling for a couple of hours, managed to train the model but am unable to generate an image because of insufficient RAM. *cries in pettiness*
Getting this errors - ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. lida 0.0.10 requires kaleido, which is not installed. llmx 0.0.15a0 requires cohere, which is not installed. llmx 0.0.15a0 requires openai, which is not installed. tensorflow-metadata 1.14.0 requires protobuf=3.20.3, but you have protobuf 4.23.4 which is incompatible. tensorflow-probability 0.22.0 requires typing-extensions
colab.research.google.com/github/huggingface/autotrain-advanced/blob/main/colabs/AutoTrain_Dreambooth.ipynb
It shows repository not found , how can I encounter this problem
@@gsomasekhar1940 its available all u need to do is refresh
Not takur, it's thakur
12 pic generation in 24 hours bruhh
Sir please make a full video on this
Varun bhaiya looks like Robert Pattinson 🗿😯 in the suit
Vengeance
😂
Right?! no ones talking about it.....
My Honest Reaction:
My honest reaction to this comment
photo studio owners be like : we never expected this AI arrow will be on us...
@@sanakchandra you're right
It's coming for everybody
@@sanakchandraI bet you can get the same quality of photos on a flagship smartphone nowadays. Even studio quality photos might cost you max 10$. Where a professional photoshoot will cost you 500$ that's when the AI comes into the play.
@@sanakchandra, keenly see, in 24hrs we all look a bit different.. in whole week as well.. so, consider the photo to be clicked on your best day while on exam you are not so bright.. then also it's same difference.. just saying, it can happen.. in camera the tech is hidden, so we assume it to be called as natural.. while, in AI, we're addressing it as tech tech tech..
even camera is a tech, i agree it's literally light mapping on it's memory/paper.. but AI is similar, kind of a artist making live portrait of you, a bit off
@@sanakchandrathings and perspectives will change
I'll say use roop extension with a good photorealistic model it gives more realistic results 👍
How to use it and where
Where to use and how to?
how to use it?
I can literally say the output face is not yours 😂.
Once you upload everything to them they can use that data for whatever they want..
@@vikramkumarp not true the model is off the shelf not an api so they don't have access to your data, don't talk when you don't know
LinkedIn and tinder profiles now 😂😂😂
Bro just ruined the entire DSLR guy's career in less than 60 seconds. 😂
In your dream 😂😂😂
@@_BSC_Cs_-BShibinyou can see it if you are not blind
2 hrs may be
@@_BSC_Cs_-BShibindon't cry end is near 🤣
@@rajwalkehe is having his own reasons because he just brought a dslr😂
Man's a 4x kaggle GM. Damn.
What
GM?
@@samarthsoni9821Grand master
@@samarthsoni9821grandmaster , kaggle is a online platform for data science and machine learning professional and students
Grandmaster
can you make a detailed video? like step by step turotial?
Exactly, even having some clue how this works, he skips like 15 steps
sooo good!!!!
Its so awesome when people collab and do wonders!!!
how to use it? i do not undertand
Guys if any of u with coding experience can simplyfy this kind of thing to a point where this can be done with a friendly UI. Trust me u r going to make millions in the next AI bubble.
What should be the project path I didn't show in there please make a clear video on this
Every time I watched his videos, I learned something new.
Great guy
After downloading the project file what should i need to do? I didn't get it please tell me
Sir please do full tutorial for beginners please 🙏
Yeah
Why did you post this now?! Just last week I needed "professional photos" for work lol
same LOL
how much it's cost you bro and for which work
Varun : Let me teach you guys AI
AI : I'm Vengeance.
2hrs vs Cost!!!
Which one you will choose!!
I'm surprised that is even a question
Truu😂
Cost
2 hrs u just need to run a script nothing else
You dont have to waste 2 hrs just 5 mins for doing your thing, then do something else in the remaining 2 hrs
Nobody can milk AI like this guy.
i wish someone could say the samee about your
@@HoTSh0t 💀😂
Good to see that the colabs is still free.
Well done on creating this app, big help for networking.
The AI shots from karma came out really good
We get 12.7 ram only in the free version of Google colab, that is not enough to generate the images. This cannot work on Google colab.
I got the same issue, let me know if you were able resolve it
@@AniketGharat pro
The output images will depend on the input image quality and the number of angles from which the image of the subject is taken. This is a very time consuming process. Or one can buy a very expensive CUDA capable graphic card and do the computation over there.
I always knew you'd look like robert Pattinson.
And the AI jus proves it. LMFAO!
Varun was looking exactly like robert Pattinson in those pictures 🔥
Such a handy free tool, appreciated!
Amazing trick, I needed this.
Very disappointed from u Varun 😔. Atleast give full tutorial 90% of the viewers couldn't able to do it and stucked at the second step😑
true
Which step you got stuck? I could it though.
@@mrinmoybanik5598After downloading the project file what should i need to do? I didn't get it please tell me
@@mrinmoybanik5598last one
@@mrinmoybanik5598i have downloaded pytorch file and thenn????
Varun, my love ❤ ur back...........
I am so happy
Ayo, Robert Pattinson is that you ?
Now, this is called True content😊
For those who don't know who Abhishek Thakur is, hes a champ on kaggle.
Its like Michael Phelps but on kaggle and kaggle is jist olympics of Data Science and ML.
First Google colab started banning accounts using free service for AI related works , specifically Stable diffusion and SDXL model is around 6.5 gb and if you feel comfortable in downloading and uploading a 6.5 gb file to Gdrive of different google accounts then its okay to follow the process as Google will ban colab services for your account as soon as you get in their system's radar... and this is an year old recipe which is outdated now
This is so cool! Can't wait to try it.
Bro I’m not kidding these AI images are fire 🔥
Should do tutorials on the other page.😊
This is great + perfect + free. Excellent YT video.
Time to catfish employers 😂
how to use it?
is there any full video?
Mostly I click like button in some videos rarely only 1 or 2 in year but you are amazing i like your every video how unique your video's are
Thanks Varun was looking for this
Did you try it? Does it work?
Meanwhile Varun looking like Robert Pattinson 😂
But this again hides the true identity of a person...if everyone starts using this, we can't really tell who that person is.
how so? It's pretty clear by the AI generated images that it's varun
But shakal toh vohi rhegi 😂 it's just like clicking it from DSLR
true identity is an interesting word, we know the face.. not the spirit n soul, so we really lack knowing who's who.. so just another level of distortion by AI
Bhai toh face kiske liye hai, the face is there right, waise bhi studio jata hu toh bhai tatti photos aati hai meri 💀
Sakshi mam apka koi photo studio hai ky?
when I run this it shows
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
tensorflow-metadata 1.14.0 requires protobuf=3.20.3, but you have protobuf 4.23.4 which is incompatible
Me too having this error
same problem bro
yeah, I tried several times but showing same code..
update the tensorflow package
@@sriansh1028 how to do that?
Thanks, finally going to change my DP after va while 😅
'Abhishek Takur' 😆 😂
ai generated video you know
@@zzyuk814 Oh bhai!!! Ye to khela hogaya...😂
Varun i training model on my images and save model file from generated folder. Can you please tell me what was the next step or what i do with this file?
bro, I am getting errors when I run the code. Can you tell me what did you do ?
If u did all this u will already be hired ngl
Thing literally made him Robert Pattinson😂
Where the generated image is saved?
it got over in 2mins and nothing was generated
same here. bro
Not Done Rightly
@@SubhajitIND then what's the right way to do it?
The 3rd cell is the main one which calls autottain's dreambooth with the parameters. It took me 100mins to run and train the model.
I can consider this trick for stock images..thanks 👍
There is alternative colab I just used to do same thing with able to download LoRA, and it is taking hardly 15 minutes
Can you help me out, doing the same thing.?
pls give me link
Brilliantly explained
Am now ready to reconnect with colleagues, thanks!
Those AI headshots are so life like, thanks karma!
I thought this will be last reel i shall watch its 4:34 am already. It's totally worth 😂🎉
I will prefer to hire someone so that he/she can grow in their business.
Thank you Abhishek 🌮
TYSM, Varun. TY!
After downloading the project file what should i do next please tell me
connect local run time and run it
Whoa the photo is so good that nobody will recognize you
Informative, well explained 👏
how to do it?
It ran out of memory while generating images
Decrease resolution then upscale later
Never have done such a work with a laptop hussh got tired seeing 😮
T4 instance (default collab instance) seems to run out of memory when trying to run the last stage of "refiner"
After downloading the project file what should i need to do? I didn't get it please tell me
Wow, It would be very useful if it generate our photos rather others 😅
Ok that image is fucking hilarious no way I'm ever using it
And those facial attributes they can misuse too.
@VarunMayya love your work but for the love of god tell your editor the sub-titles are competing with Ranveer Sigh - they are so everywhere so hard to keep track,
Vengeance Varun 🦇
I'd go to the studio and get myself clicked instead of putting 2 hrs. Then run some errands and come back home with the edited photos already in my email.
Abhishek Takoor 😅😅😅
Robert Pattinson from The Batman ❤
train has failed due to an exception: RuntimeError: element 0 of tensors does not require grad and does not have a grad_fn
Abhishek thakur is one of the best ai expert of india and world
Thank you Varun I will definitely use it ❤
did it work?
@@Amber-zp1zr I haven't till now but will do surely and let you know 😄
U did thakur dirty😂
Correction: One of the best science fiction movie.
kudos ... great content on ai
This code to prompt did not work due to Indentation Error. Kindly support
Robert Mayya
yea those pics actually look like Robert Pattinson.
Abhishek is OG in AI/ML space!
Still the photoshoot will give a much more creative and natural looking photo than those sasta AI tricks.
If possible I'll just create an app using the model code and share it for free. 👍
After downloading the project file what should i need to do? I didn't get it please tell me
It's just a bit weird to see how people are moving away from genuineness....
Great information sir 👍 👏
Yes. But the uncanny valley issue is there.
Save kar le ta hoon future mein kaam aayega😂
He is a vampire ( Pattinson).
Bruh, after struggling for a couple of hours, managed to train the model but am unable to generate an image because of insufficient RAM. *cries in pettiness*
Good Abhishek thakur is awesome
I would say natural photo is all you need, which describe you what you are rather AI generated. faked emotion on photo people have to think 100 times
I got the file, I didn't understand where you enter the key and the prompt after that.
How did you get the file? After the running, I don't see a file anywhere
Ai will save a lot of money in marketing
please don't exaggerate accent , that's not takur , it is thakur
YES pronunciation of this word should sailent the word h but if it's NAME 🙅♀️❌
Getting this errors -
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
lida 0.0.10 requires kaleido, which is not installed.
llmx 0.0.15a0 requires cohere, which is not installed.
llmx 0.0.15a0 requires openai, which is not installed.
tensorflow-metadata 1.14.0 requires protobuf=3.20.3, but you have protobuf 4.23.4 which is incompatible.
tensorflow-probability 0.22.0 requires typing-extensions
same error
Abhishek Takur is not the person who came up with autotrain it a library developed by Huggingface
@Varunmayya Its not working. Can you please re check it or just upload a full video on it
Now what will happen to photographers 😢