---
tags:
- text-to-image
- flux
- lora
- diffusers
- template:sd-lora
- fluxgym
widget:
-   output:
        url: sample/ldlaughingmemeface_000450_00_20241001223402.png
    text: LDME Young adult male at a party, mid-laugh with eyebrows raised in mock
        surprise. He's holding a red solo cup and wearing a graphic t-shirt
-   output:
        url: sample/ldlaughingmemeface_000450_01_20241001223612.png
    text: LDME Close-up of a smirking businessman with arched eyebrows, squinting
        eyes, and a slightly open mouth. He's holding a coffee mug in an office setting
-   output:
        url: sample/ldlaughingmemeface_000450_02_20241001223822.png
    text: LDME middle-aged woman in a casual shirt, raising his eyebrows skeptically
        while holding a martini
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: LDME
license: other
license_name: flux-1-dev-non-commercial-license
license_link: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md
---

# LDLaughingMemeFace

A Flux LoRA trained on a local computer with [Fluxgym](https://github.com/cocktailpeanut/fluxgym)

<Gallery />

## Trigger words

You should use `LDME` to trigger the image generation.

How to Use This LoRA Model

Describing the Expression

To get the best results from this LoRA model, use descriptive words that capture the essence of the facial expressions you want to generate. Consider phrases like:

"exaggerated expression"
"raised eyebrows"
"as if mocking"
"eyebrows lifted and smirking"

These descriptors will help the model focus on creating the unique, expressive faces you're aiming for.

Adjusting LoRA Strength

Strength 1: At full strength, the LoRA model tends to generate images of the same character when you prompt with "a man." This can be useful if you want consistency in character appearance.

Dialing Down Strength: If you wish to create different characters or add variety, reduce the strength of the LoRA. A strength setting around 0.7 has been effective for me, but I encourage experimenting with different levels to achieve your desired outcome.

## Download model and use it with ComfyUI, AUTOMATIC1111, SD.Next, Invoke AI, Forge, etc.

Weights for this model are available in Safetensors format.