Lostreak Fusion style and character Lora all in one [PonyXL]

Details

Download Files

About this version

Here is the prompt data if it helps:

"l0streakchick3n": 84, "d0gg0u": 70, "1boy": 61, "black background": 7, "white background": 6, "simple background": 16, "half-closed eyes": 1, "suprised look": 1, "pointing at viewer": 2, "tongue": 1, "open mouth": 8, "teeth": 5, "smile": 11, "sunglasses": 1, "eyewear on head": 1, "panth3rba0": 57, "1girl": 58, "holding butt": 1, "holding person": 1, "back": 5, "rear view": 4, "hands": 41, "angry": 2, "angry eyes": 1, "screaming": 1, "chinese text": 3, "smug expression": 4, "tiles": 6, "bathoom": 3, "upper body": 6, "bathroom": 3, "suggestive pose": 2, "closed eyes": 7, "arms behind back": 1, "sink": 2, "br0therli0n": 15, "restaurant": 11, "booth seating": 11, "scared": 6, "worried": 8, "serious face": 4, "disgust": 2, "squinting": 7, "sad": 8, "blush": 13, "arms around shoulders": 7, "close-up": 9, "knees to chest": 3, "looking at phone": 7, "holding phone": 18, "looking at another": 7, "standing": 9, "smug": 1, "laughing": 10, "living room": 12, "on couch": 10, "couch": 16, "holding": 9, "sitting": 11, "crying": 2, "leaning on person": 1, "holding mouth open": 3, "sitting on person": 1, "phone screen": 4, "grin": 3, "photo \\(object\\)": 2, "light smile": 1, "bedroom": 8, "window": 1, "2boys": 12, "poster \\(object\\)": 10, "eating": 5, "pondering": 1, "hand over shoulder": 3, "walking": 3, "holding money": 1, "park": 3, "fence": 1, "tree": 1, "from side": 1, "crossed legs": 1, "table": 7, "holding can": 1, "wide eyed": 4, "looking at viewer": 2, "holding bag": 2, "chips \\(food\\)": 2, "lotus position": 2, "looking down": 2, "at night": 3, "refrigerator": 2, "crossed arms": 3, "text": 3, "gaming headset": 1, "door": 8, "open door": 2, "playing games": 1, "gaming chair": 1, "looking back": 1, "looking back at another": 5, "blanket": 1, "gym": 5, "boxing gloves": 2, "kitchen": 3, "holding boxcutter": 1, "kitchen counter": 2, "dialated pupils": 1, "suprised eyes": 1, "holding pillow": 1, "hallway": 2, "dim lighting": 2, "tile floor": 1, "smiling at another": 2, "on bed": 3, "pointing at another": 2, "pointing at head": 1, "constricted pupils": 2, "extended arms": 5, "happy": 2, "panther_ba0": 1, "dog_g0u": 1, "from behind": 1, "sitting on bed": 1, "walking away": 1, "lying on bed": 2, "bed": 2, "sunrise": 2, "fleeing": 1, "open smile": 1, "holding box": 1, "blank expression": 3, "sitting on sofa": 2, "livingroom": 2, "sunset": 5, "opening door": 1, "box": 1, "sky": 4, "cloud": 4, "city": 2, "dusk": 2, "lobster": 3, "pancake": 3, "stuffing": 3, "holding bottle": 2, "vending machine": 2, "park bench": 1, "hat": 1

Model description

This Lora is inspired by a web animated series by 星有野 on bilibili:https://space.bilibili.com/627888730. It took me quite a while to tag all images manually but I think it turned out well enough. Comment if you want me to include other characters from the series or just general improvements. Post your creations I would like to see what you make with this model. Happy generating!

V2 Multipurpose

For this version I completely redid the dataset and everything. This version can do a lot of things for example if you like the style but don't want the characters just use the lora without the prompt, if you get anthro humans just put the word: anthro in the negative, see the images for reference.

Now for the characters I added two more characters the Rooster and the Bunny Girl.

Usage:

I use ADetailer and regional prompter for some of the images.

Strength: 0.7-1 that's what I normally use but feel free to experiment.

For example if I want to make an image of Panther Bao I would add the prompts like this:

panth3rba0, 1girl, then describe what you want, if the effect is not strong enough add l0streakchick3n, anthro at the beginning.

Trigger words:

Panther Bao--- panth3rba0

Dog Gou--- d0gg0u

Brother Lion--- br0therli0n

Rooster--- r00ster

Bunny Girl--- b0nneygir1

l0streakchick3n, anthro--- this changes the image up a bit not really necessary for prompt but interesting different results, if you don't want the characters put this in the negative prompt

NO PROMPT--- with no prompt you just get the style of the model and not the characters

Examples:

Different checkpoints:

Lora on/off:

V1

Usage:

This version is trained on 3 characters:

Panther Bao

Dog Gou

Brother Lion

Trigger words:

Panther Bao--- panth3rba0

Dog Gou--- d0gg0u

Brother Lion--- br0therli0n

l0streakchick3n--- this changes the image up a bit not really necessary for prompt

You can also add 1boy or 1girl depending on the character you want.

For strength I usually use 0.7-1.

As always have fun.

If you want you can support me here all donations are greatly appreciated:https://ko-fi.com/n3o_wav3

Images made by this model

Sort by

No Images Found.

Model Type	LORA
Base Model	Pony
Published	7/12/2024
Trained Words	l0streakchick3n panth3rba0 d0gg0u br0therli0n