2000 Ads

✨ Features

• Generate unique advertisements in a specific style using LoRA weights.
• Compatible with multiple UIs such as AUTOMATIC1111, Comfy UI, SD.Next, Invoke.
• Can be used with the 🧨 diffusers library for programmatic image generation.

📦 Installation

Download the Model

• LoRA: Download 2000_ads.safetensors here 💾.
  • Place it in your models/Lora folder.
  • On AUTOMATIC1111, load the LoRA by adding <lora:2000_ads:1> to your prompt. On ComfyUI, just load it as a regular LoRA.
• Embeddings: Download 2000_ads_emb.safetensors here 💾.
  • Place it in your embeddings folder.
  • Use it by adding 2000_ads_emb to your prompt. For example, an ad in the style of 2000_ads_emb (you need both the LoRA and the embeddings as they were trained together for this LoRA).

💻 Usage Examples

Use it with the 🧨 diffusers library

from diffusers import AutoPipelineForText2Image
import torch
from huggingface_hub import hf_hub_download
from safetensors.torch import load_file

pipeline = AutoPipelineForText2Image.from_pretrained('stabilityai/stable-diffusion-xl-base-1.0', torch_dtype=torch.float16).to('cuda')
pipeline.load_lora_weights('LinoyTsaban/2000_ads', weight_name='pytorch_lora_weights.safetensors')
embedding_path = hf_hub_download(repo_id='LinoyTsaban/2000_ads', filename='2000_ads_emb.safetensors', repo_type="model")
state_dict = load_file(embedding_path)
pipeline.load_textual_inversion(state_dict["clip_l"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder, tokenizer=pipeline.tokenizer)
pipeline.load_textual_inversion(state_dict["clip_g"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder_2, tokenizer=pipeline.tokenizer_2)

image = pipeline('<s0><s1> ad of a llama wearing headphones').images[0]

For more details, including weighting, merging and fusing LoRAs, check the documentation on loading LoRAs in diffusers.

📚 Documentation

Trigger Words

To trigger image generation of the trained concept (or concepts), replace each concept identifier in your prompt with the new inserted tokens: To trigger concept TOK → use <s0><s1> in your prompt.

🔧 Technical Details

• All Files & versions: /LinoyTsaban/2000_ads/tree/main.
• Training Script: The weights were trained using 🧨 diffusers Advanced Dreambooth Training Script.
• LoRA for the text encoder: Enabled: False.
• Pivotal tuning: Enabled: True.
• Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.

📄 License

This model is licensed under the openrail++ license.