Excellent tutorial! Congratulations, David and thank you! I would like to ask a question. It is possible to include more than one training session in the same session. For example, aiglasse + aiclothing? How could I do that. thank you! :)
Ahhhh great now I see clearly how this works, quite involved but very doable thanks for highlighting. I’m wondering if training a life coach clients face and likeness how realistic the outputs are for her viewers to think they really are actual photographs
Just got paid doing this on my own and came for extra tips for success, for those saying it doesn’t work you are wrong I’ve had a business for a year and my first sale was using ai to create logos and cover art.
As Usual David you giving away a Million Dollar Skill the only people that would say it’s not is ignorant to the Advertising Space and don’t understand the power of AI 4 and 5 figure skills learned in a 20 min Video I made my son(which has autism )learn this skill now he’s able to charge influencers and other small companies a few thousand dollars for two hours worth of work and I am to, thank you David
This worked wonders! I used my own images and it generated amazing professional images of myself which I can really use on LinkedIn! Will definitely work on building a business with this!
David, very good video. What takes the most effort is finding a business that has about 50 photos of the same product to have a good dataset. Any recommendations?
Hi. Can I use a LoRA trained on Replicate with other accounts? In other words, can I train my LoRA with my face or a consistent character and share it for others to use with their accounts? If so, how does the process work? Thanks a lot
Thanks for this video. How would you do a LoRA with the Flex Pro version? Wouldn't you want to use pro vs dev for professional and commercial uses? Much appreciated.
The licence is only limited to the usage of the model itself not the actual images generated. For example if you wanted to make a commercial app that uses the dev model that wouldn't align with the dev's licence's rules. This is not the same for generated images
it seemed to work (getting charged) but I have nothing in the Model card in Hugging face. And when I run the model in Replicate, the generated images have nothing to do with the images I trained it with.
Do you know how to let someone upload photos and train on their images so they van do it themselves if you provide an app that can upload and create images and train etc
I must be doing something wrong. I got a few images with my trained model (of myself) but now I am getting images without any hint of my uploaded images/trained model. Is there a way that I am not using my own trained model?
You would be surprised what people pay big money for. That being said you do need to get your hands dirty and get some level of expertise if you want to get paid
Logical thinking. However logic isn’t always applicable in the real world. You can spend 20 minutes and learn how to run Tik Tok ads, but there are still many ad agencies out there.
People already get these services easily on Fiverr since last year and even earlier.... Dozens of AI workers who do exactly that, but it is competitive and no they're not getting rich from that 😅
When you were trying out the fine tuned model in replicate, did u have to deploy it in a server to try it? Per replicate pricing it says.. "Unlike public models, you’ll pay for setup and idle time in addition to the time it spends processing your requests." THank you
If you train model in 1000 steps, is there a way to save file every 200 steps? And can i write captions like this "man, upper body, blue eyes, hat" or do I have to write it in sentences like "picture of a man with blue eyes wearing a hat" ??
Do you need to train the model in how to draw a person? The majority of the results has no leg, or two heads, bad proportions, etc. I gave it 60 pictures.
🔥 Start making money with AI: www.skool.com/new-society
I love your channel 🎉thank you for sharing 😊
Great tutorial! It was super helpful for understanding how to fine-tune FLUX.1 with custom LoRA. Thanks for the clear explanations! 🌟
Awesome video guide! Super happy that you like the Flux Trainer
David, man, you made a great video and explainer for this new technique! But more then that, a great strategy for selling the service. Grazie!!
So excited! Damn good tutorial and a genuinely excellent video. I do believe there's a few dollars to be made out there!!!
You got to respect the hustle. Subbed.
Thanks man! Love what you are doing!
A video with substance, Appreciate it.
Extremely valuable and easy to follow, thank you.
Amazing video! Lots of value!!
Brilliant walk through. Successful at the process, but I def could use some tips on prompting. Exciting about joining!
Excellent tutorial! Congratulations, David and thank you! I would like to ask a question. It is possible to include more than one training session in the same session. For example, aiglasse + aiclothing? How could I do that. thank you! :)
I respect you, sir :)
garbage in garbage out, well said bro! thanks for tutorial !😀
Ahhhh great now I see clearly how this works, quite involved but very doable thanks for highlighting. I’m wondering if training a life coach clients face and likeness how realistic the outputs are for her viewers to think they really are actual photographs
Wow
Excellent video bro🤝
Great tutorial ! thank you very much !
19:48 That's how "getting told off by means of looking" looks like 😂
Just got paid doing this on my own and came for extra tips for success, for those saying it doesn’t work you are wrong I’ve had a business for a year and my first sale was using ai to create logos and cover art.
Yo bro, ive trained my own lora as well. just need quick tip on how you got paid for this?
As Usual David you giving away a Million Dollar Skill the only people that would say it’s not is ignorant to the Advertising Space and don’t understand the power of AI 4 and 5 figure skills learned in a 20 min Video I made my son(which has autism )learn this skill now he’s able to charge influencers and other small companies a few thousand dollars for two hours worth of work and I am to, thank you David
Thank you!
Amazing! Thanks!
This worked wonders! I used my own images and it generated amazing professional images of myself which I can really use on LinkedIn! Will definitely work on building a business with this!
I know I can skip the hugging face part and train the model just in Replicate. What’s the difference and what value hugging face adds to the process?
Díky!
in Think Deep we love u
Monthly is quite costly and if its a one time membership for 20dollars than better and it should be based on region
David, very good video. What takes the most effort is finding a business that has about 50 photos of the same product to have a good dataset. Any recommendations?
thanks for the video! what is the purpose of uploading it to huggingface? i didnt add a repo for hf and it worked out too
The license does not allow you to make money with the dev model, and if you train a Lora, it has the same license as the model you are training for...
just change to shell flux it can comercil
Hi. Can I use a LoRA trained on Replicate with other accounts? In other words, can I train my LoRA with my face or a consistent character and share it for others to use with their accounts? If so, how does the process work? Thanks a lot
Thanks for this video. How would you do a LoRA with the Flex Pro version? Wouldn't you want to use pro vs dev for professional and commercial uses? Much appreciated.
The licence is only limited to the usage of the model itself not the actual images generated. For example if you wanted to make a commercial app that uses the dev model that wouldn't align with the dev's licence's rules. This is not the same for generated images
it seemed to work (getting charged) but I have nothing in the Model card in Hugging face. And when I run the model in Replicate, the generated images have nothing to do with the images I trained it with.
Is it safe with bank info and all that? I would love it if it had a paypal option.
Is this LORA actually free or do they bave an upper cap on kn usage ?
Do you know how to let someone upload photos and train on their images so they van do it themselves if you provide an app that can upload and create images and train etc
How much money do you spend to train a model? I.e. what is the cost of training for all accounts and services needed?
bro i want to make my own lora, so how can i put it public, is there any more ways? 8:59
Is it posibble to upload the model directly to replicate instead of hosting it in huggingFace?
I must be doing something wrong. I got a few images with my trained model (of myself) but now I am getting images without any hint of my uploaded images/trained model. Is there a way that I am not using my own trained model?
how do you make face video? is it some program or AI? I see how cut off it comes because it's cut to what you do, how do you make this? its amazing
Is there a way to use this model to add the object to an already existing image?
But why company will pay us to do this when they can just have someone in the company doing all the time?😅 4:28
Can you train one with Lebron James and another with some Nike models and generate cheap Ads?
Would like a Flux Pro version of this trainer :)
I don't think anyone is going to pay big money for a skill that can be learned from a 20 min YT video.
Of course they will - anything that saves time = money
You would be surprised what people pay big money for. That being said you do need to get your hands dirty and get some level of expertise if you want to get paid
Logical thinking.
However logic isn’t always applicable in the real world.
You can spend 20 minutes and learn how to run Tik Tok ads, but there are still many ad agencies out there.
You get average users then you get advanced user that put 100s of hours improving their skill. They can make money
People already get these services easily on Fiverr since last year and even earlier.... Dozens of AI workers who do exactly that, but it is competitive and no they're not getting rich from that 😅
When you were trying out the fine tuned model in replicate, did u have to deploy it in a server to try it? Per replicate pricing it says.. "Unlike public models, you’ll pay for setup and idle time in addition to the time it spends processing your requests." THank you
thx
If you train model in 1000 steps, is there a way to save file every 200 steps?
And can i write captions like this "man, upper body, blue eyes, hat" or do I have to write it in sentences like "picture of a man with blue eyes wearing a hat" ??
Do you need to train the model in how to draw a person? The majority of the results has no leg, or two heads, bad proportions, etc. I gave it 60 pictures.
did you remove the RAG videos?
Can i use this LoRa on flux-nf4 on Forge?
Can I train on my Apple M3 Max machine?
can we use flux for commercial use? and also the lora for commercial?
you can help me i have problem with folder photo
Tyyyyyyyst
hi. i thought price increase was at 700 members. any way to get the old price?
your links dont work to skool
use confyUI to use your model it cost 0 if you have a good video card like rtx 4070 and more
Can you train it for infographics?
its free ?
20:00
click bait once more, paid services, total BS
haha.....If you have the hardware, you can train it yourself.
Is there a way to fine tune for free?
@@hazemkhaled2714 maybe if u had a few 4090s sitting around
kahma AI fixes this. Create AI photos resembling you.
No matter what I did, my photo could not be uploaded and gave an error
This field is required
same
try training with the api, i had the same problem and it worked using the api
@@estefanoandres How do you give me the link? I am an amateur
i dont know why but i couldnt watch the video or focus because his headphones are falling off his head.