YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/model-cards#model-card-metadata)

Installtion

sudo apt-get update && sudo apt-get install git-lfs ffmpeg cbm

conda create --name py310 python=3.10
conda activate py310
pip install ipykernel
python -m ipykernel install --user --name py310 --display-name "py310"

pip install diffusers torch torchvision moviepy==1.0.3 omegaconf compel transformers peft safetensors datasets tabulate
git clone https://huggingface.co/svjack/Star_Rail_Tribbie_Lora

Inference

from compel import Compel, ReturnedEmbeddingsType
from diffusers import DiffusionPipeline, StableDiffusionXLPipeline
import torch

# Initialize the DiffusionPipeline with SDXL
pipeline = StableDiffusionXLPipeline.from_single_file(
    "Star_Rail_Tribbie_Lora/waiNSFWIllustrious_v110.safetensors",
    torch_dtype=torch.float16
)
pipeline.enable_model_cpu_offload()

# Step 4: Load the LoRA adapter
pipeline.load_lora_weights("Star_Rail_Tribbie_Lora/tribbie.il.safetensors")

# Initialize Compel for SDXL
compel = Compel(
    tokenizer=[pipeline.tokenizer, pipeline.tokenizer_2],
    text_encoder=[pipeline.text_encoder, pipeline.text_encoder_2],
    returned_embeddings_type=ReturnedEmbeddingsType.PENULTIMATE_HIDDEN_STATES_NON_NORMALIZED,
    requires_pooled=[False, True],
    truncate_long_prompts=False,
)

prompt = "1girl ,detailed background, posing, <lora:tribbie.il:1> tribbie, cross-shaped_pupils , one eye closed, missile,"
prompt = "1girl, detailed background, posing,  <lora:tribbie.il:1>, tribbie, cross-shaped_pupils, one eye closed, holding a cup of coffee, relaxed atmosphere, cozy cafe setting, warm lighting, steam rising from the coffee cup, soft smile, casual outfit, sitting by a window, outside view of a peaceful street, plants in the background, soft focus on the background to emphasize the character."
prompt = "detailed background, shiny skin, posing, <lora:tribbie.il:1> tribbie, 3girls, eye patch, hair over eye,"

# Generate conditioning tensors
conditioning, pooled = compel(prompt)

# Generate the image
image = pipeline(
    prompt_embeds=conditioning,
    pooled_prompt_embeds=pooled,
    num_inference_steps=30
).images[0]

# Save the image
image.save("tribbie_image.jpg")

# Display the image
from IPython import display
display.Image("tribbie_image.jpg", width=512, height=512)

Demo

  • 1girl ,detailed background, posing, lora:tribbie.il:1 tribbie, cross-shaped_pupils , one eye closed, missile,

image/jpeg

image/jpeg

  • 1girl, detailed background, posing, lora:tribbie.il:1, tribbie, cross-shaped_pupils, one eye closed, holding a cup of coffee, relaxed atmosphere, cozy cafe setting, warm lighting, steam rising from the coffee cup, soft smile, casual outfit, sitting by a window, outside view of a peaceful street, plants in the background, soft focus on the background to emphasize the character.

image/jpeg

image/jpeg

  • detailed background, shiny skin, posing, lora:tribbie.il:1 tribbie, 3girls, eye patch, hair over eye,

image/jpeg

python produce_tribbie_dataset.py
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model has no library tag.