Instructions to use Efficient-Large-Model/SANA1.5_4.8B_1024px with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Sana
How to use Efficient-Large-Model/SANA1.5_4.8B_1024px with Sana:
# Load the model and infer image from text import torch from app.sana_pipeline import SanaPipeline from torchvision.utils import save_image sana = SanaPipeline("configs/sana_config/1024ms/Sana_1600M_img1024.yaml") sana.from_pretrained("hf://Efficient-Large-Model/SANA1.5_4.8B_1024px") image = sana( prompt='a cyberpunk cat with a neon sign that says "Sana"', height=1024, width=1024, guidance_scale=5.0, pag_guidance_scale=2.0, num_inference_steps=18, ) - Inference
- Notebooks
- Google Colab
- Kaggle
Inference
#1
by poluramus - opened
Hello!
First, thank you for this release ๐
I tried the diffusers code from https://huggingface.co/Efficient-Large-Model/SANA1.5_4.8B_1024px_diffusers, but the outputs are great. Itโs probably due to the PR not being merged?
CMD:
Some weights of the model checkpoint at /home/keven/.cache/huggingface/hub/models--Efficient-Large-Model--SANA1.5_4.8B_1024px_diffusers/snapshots/70974747bbd7a8abbd7b3d954c2e6f4b0d5750ad/transformer were not used when initializing SanaTransformer2DModel:
['transformer_blocks.18.attn2.norm_q.weight, transformer_blocks.3.attn1.norm_k.weight, transformer_blocks.49.attn1.norm_k.weight....
And how should we use the model in this repo ?
Thank you :-)
poluramus changed discussion status to closed