SDXL — Product Photography LoRA
Fine-tuned on 20 studio sneaker images using DreamBooth LoRA on Kaggle T4 (15GB).
Training: 800 steps, ~2.5 hours. Trigger word: PRDCT
Usage
from diffusers import StableDiffusionXLPipeline
import torch
pipe = StableDiffusionXLPipeline.from_pretrained(
"stabilityai/stable-diffusion-xl-base-1.0",
torch_dtype=torch.float16
)
pipe.load_lora_weights("Shoaibism/sdxl-product-photography-lora")
pipe.enable_model_cpu_offload()
image = pipe(
"a photo of a PRDCT white sneaker, studio background, professional product photography, 8k",
height=1024, width=1024, num_inference_steps=25, guidance_scale=7.5,
).images[0]
image.save("result.png")
Before / After
Portfolio Samples
Author
Muhammad S. — AI/ML Engineer Upwork
- Downloads last month
- 78
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for Shoaibism/sdxl-product-photography-lora
Base model
stabilityai/stable-diffusion-xl-base-1.0