Here is my realistic Mona Lisa Lora that you can use with FLUX. She was trained on 91 images and is quite expressive as well because of her dataset.
Make sure to put the Lora strength on 1.5 for the best results.
The triggerword is moonal1saa
Example prompts to try:
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> moonal1saa posing for a fashion campaign wearing an elegant and ornate red blazer and black sunglasses
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> a extreme closeup portrait of moonal1saa screaming and taking a selfie with a llama in peru, she is wearing a peruvian cotton jacket and beanie
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> a closeup portrait of moonal1saa smiling, she is a female wizard wearing colorful robes, she is holding a cat
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> a extreme closeup portrait of moonal1saa wearing a elegant white jacket with diamonds, standing on a rooftop with the amalfi coast of Italy in the background, she is holding a glass of red wine
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> a portrait of moonal1saa wearing red lipstick and a red cardigan, holding red lipstick in her right hand
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> moonal1saa fashion model posing, wearing a black t-shirt on a fashion runway
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> closeup of moonal1saa fashion model posing, wearing a black renaissance gown with gold ornate detailing, high fashion, gucci, on a fashion runway
<lora:real_mona_lisa_flux_lora_triggerword_moonal1saa:1.5> extreme closeup of moonal1saa fashion model posing, wearing a black renaissance bomber jacket with gold ornate detailing, high fashion, gucci, on a fashion runway
Enjoy!
Description
This is the final trained version of the realistic Mona Lisa Lora.
FAQ
Comments (7)
where did you get 91 images?
I used Flux PULID to build the dataset to obtain 91 images since it allows you to create a consistent character from a single image so I used the original painting and it gave me a realistic version. Then I built a modified dataset with different clothing, lighting conditions and poses. I also used Comfy UI's Advanced live portrait feature to generate a wide range of facial expressions to make her expressive.
@Trav1 Can you tell me how many epochs / total steps you trained her? My lora is about 38 mb in size and a bit lower quality. I am using flux gym for training.
Great job bro. In the end she turned out to be much prettier than we had always thought xd
Thank you JabGal, yeah I think she turned out to be quite the hotty hahaha.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.

















