Note: This step is irreversible for that specific model repository but is mandatory for ethical AI usage.
Whether you are a developer looking to integrate AI into your application or an enthusiast setting up a local generator, knowing how to properly download runwayml/stable-diffusion-v1-5 from Hugging Face is an essential skill.
The model is primarily hosted on Hugging Face. Access requires a registered account and acceptance of the model license terms.
Downloading runwayml/stable-diffusion-v1-5 from Hugging Face is straightforward but requires: runwayml stable-diffusion-v1-5 huggingface download
The most seamless way to use the model in Python is via the Hugging Face diffusers library. This method handles the download, model architecture, and safety checkers automatically.
Even if the download succeeds, loading the model might crash your GPU.
pixels) from text prompts. It was initialized from the Stable Diffusion v1.2 weights and further trained for 595k steps on the "laion-aesthetics v2 5+" dataset. Note: This step is irreversible for that specific
Since the repository is now "gated," you cannot simply copy-paste a download link. You must authenticate your environment with Hugging Face.
pipe = StableDiffusionPipeline.from_pretrained("./stable-diffusion-v1-5", torch_dtype=torch.float16) pipe.to("cuda") image = pipe("a photograph of an astronaut riding a horse").images[0] image.save("test_output.png") print("Model works correctly.")
from huggingface_hub import snapshot_download Access requires a registered account and acceptance of
# Save the image image.save("astronaut_rides_horse.png")
If you are using a UI like or ComfyUI that requires the .safetensors or .ckpt file specifically, you can download the raw weights manually.