AveroLabs/Avero-Image-Coming-Soon
0
Avero Image - Fine-tuned Stable Diffusion v1.5
This model is a fine-tuned version of Stable Diffusion v1.5 trained on a diverse dataset spanning fashion, nature, city, portrait, food, and animal categories.
Training Details
- Base Model: runwayml/stable-diffusion-v1-5
- Training Steps: 8,000
- Resolution: 512x512
- Optimizer: 8-bit Adam
- Learning Rate: 5e-6 with cosine schedule (500 warmup steps)
- Batch Size: 1 with gradient accumulation 4
- Hardware: NVIDIA T4
- Precision: Mixed FP16
Categories
- Fashion (16 queries): Runway, editorial, street style, couture
- Nature (16 queries): Mountains, ocean, forest, landscapes
- City (15 queries): Urban, architecture, skyline, streets
- Portrait (12 queries): Bokeh, studio, artistic, environmental
- Food (10 queries): Gourmet, cuisine, desserts, plating
- Animals (10 queries): Wildlife, pets, nature scenes
Usage
from diffusers import StableDiffusionPipeline
import torch
pipe = StableDiffusionPipeline.from_pretrained(
"itriedcoding/avero-image",
torch_dtype=torch.float16,
).to("cuda")
image = pipe("a beautiful landscape", num_inference_steps=50).images[0]
image.save("output.png")Example Prompts
- "fashion runway gown"
- "mountain sunset landscape"
- "cyberpunk city night"
- "portrait bokeh lens"
- "gourmet plated dish"
- "wildlife lion savanna"
