Latest page update: Updated recommended workflow again to a version 2, which is slightly even better.
My recommended local ComfyUI workflow for best results:
Trigger (must include in prompt): early 2010s snapshot photo captured with a phone and uploaded to facebook, featuring dynamic natural lighting, and a neutral white color balance with washed out colors, picturing
Please note that the online generator will NOT give you as good results as running the model locally in ComfyUI, because the online generator does not have the recommended sampler and scheduler and other customization options provided by my recommended local ComfyUI workflow!
Donations: If you want to support me financially, you can donate here: https://ko-fi.com/aicharacters
FINAL FINAL version notes:
I guess I lied and I am still updating this model after all
noticeably improved style likeness again, improved style consistency massively
new recommended inference settings
LoRa description:
This LoRa model was designed to generate images that look more "real" and less "artificially AI-generated by FLUX". It achieves this by making the natural and artificial lighting look more real and making the bokeh less strong. It also adds details to every part of the image, including skin, eyes, hair, and foliage. It also aims to reduce somewhat the common "FLUX chin and skin" and other such issues.
Training Workflow:
18 images at 1 repeat. ChatGPT generated captions. Trained on the default FLUX.1-dev model.
Disclaimer:
I do not claim to be the current owner, original creator of, person or legal representative of a person depicted in, any of the concepts that my models, including this one, aim to emulate nor the training data used to train this model. I do not claim that any of my models have received an endorsement by those respective individuals. I also do not claim that the emulation attempted by my models represents a 100% accurate depiction of the original concept in question nor do I claim that the quality of my emulation reaches the same quality of the original concept. All credit goes to the respective current owners, original creators, or people and I encourage you to support them in any way you can.
If you are the current owner, original creator of, person or legal representative of a person depicted in, any of the concepts that any of my models aims to emulate and would like that specific model to be removed, then please leave me a private message here or on Reddit with proof of authenticity of your identity and claim and I will remove it.
Additionally, I do not endorse my models being used to violate the law or enact still legal, but immoral acts such as the spread of misinformation using deep fakes.
Description
FAQ
Comments (27)
Okay so when is the FINAL, FINAL, FINAL version coming?
just after the FINAL, FINAL REAL version i guess ...
you're now obligated to release a FINAL_FINAL_v2_FOR REAL THIS TIME_Final
'tis da year 2069, sus. deadass finna drop fr boujee v420. [FINAL FINAL ON GOD NO CAP SWEAR ON ME MUM'S HEAD WORD ON MY DADDY'S GRAVE NO KIZZY BIG MAN TING ONLY KEEPIN' IT A BUCK FIFTY FAM]
Bet, big Ws in chat dis shit lit im bussin. high key a real CEO aura farming dank drip.
I would love to use this LORA, it looks great. But it does not work in InvokeAI. The error is:
RuntimeError: The base model quantization format (likely bitsandbytes) is not compatible with DoRA patches.
All other Flux LORAs work fine though. There is something specific about this one.
Sorry to hear that, but there is nothing I can do about that. Thats an issue with Invoke. I cannot change my training config.
Do you use fp8_scaled? If so try and use fp8 (not scaled), e.g. from Kijaj.
great work dude. the people in your reddit threads are such assholes always feel bad for you
Nice work. sorry for asking I'm pretty noob and new. Can i use this with the new flux kontext ?
Technically yes but realistically no. I tried it and the output is bad. This deems tl be true for all of my LoRa's.
I retrained one of my LoRa's for Kontext specifically and resutrs were much better. So i will provide Kontext-specific versions for all models soon, including this one.
Looks great! how does it compare to chroma? I had good results with chroma and trying to build aichroma.site for people to try!
I cannot train chroma until kohya implements it in their trainer as other trainers that already have chroma support dont work as well for me.
Thanks for sharing the workflow! It makes a huge difference in the results!
I use the exact same thing in your workflow, such as checkpoints and clips, but only images with full noise are generated.
Do you know what the cause is?
Have you tried using it with your own workflow which you know works? Otherwise I dont know what the issue is. It works fine on my end and other peoples ends. Only issue ive heard reported is with the precision in other programs like Invoke or something.
same here
@iakkm @Shuntea Do you guys use something other than normal ComfyUI (e.g. InvokeAI or something else)? Have you tried your own workflows which you know work?
with Forge only work the v12, with other version always noise is generated.
Is this actually a "DORA", as the description states, or is it a "LORA" ? Was searching for loras (via filter) and was confused why I cannot find this page - well, that is why.
the samplers used in your workflow aren't available in ComfyUI. where did you get them?
@bbmin7b5 Sorry, but sometimes I wonder why I even bother typing all this stuff out when nobody ever reads my model descriptions or workflow descriptions. I spend so much time and effort on this stuff and its all for naught. The amount of questions I get that are already answered by my description is too fucking many.
I have a note node in the workflow that answers this already: https://imgur.com/a/9zeigb7
Hi, one question, only version 12 works with Forge, all the others just make noise, is there any solution or just comfy?
Seems that Forge doesnt work with DoRa's. I cant fix that. Thats something Forge must fix. But I may at some point retrain to a normal LoRa again. I managed to get a similar config going that works almost as well as my DoRa's but as a LoRa. But I dont have the money to do all the retraining atm.
AI_Characters Thanks for the reply, Forge probably still has some limitations.
Thanks bro, trying to get this to work with flux fill (kontext gives poor results compared to that) for inpainting/outpainting, any tips on modifying the workflow to suit?
I get an error saying: Missing Node Types
Scheduler Selector (Comfy) (Image Saver)
Me too, what is going on? How to fix it?
Details
Files
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
Mirrors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
Flux_AmateurSnapshotPhotoStyle.safetensors
amateur-photo.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
amateur_snapshot_photo_v16.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
amateurSnapshotPhotoSTYLE_v160FINALFINAL.safetensors
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.








