Stable Diffusion 1.5
#46 en Texte-vers-imageStability Ai · v1.5 · depuis Oktober 2022 · 3× · vu le 30 juin 2026
1
Momentum
Stable Diffusion 1.5 is an open-source latent diffusion model (LDM) released by RunwayML / Stability AI in October 2022. It generates 512×512-pixel images from text prompts using an 860M-parameter UNet paired with a frozen CLIP ViT-L/14 text encoder. The model was trained on the LAION-Aesthetics v2 5+ dataset for 595,000 steps at 512×512 resolution and is licensed under CreativeML OpenRAIL-M, permitting free commercial use. It remains the most widely used base model for community fine-tuning (LoRA, DreamBooth, Textual Inversion) and continues to be supported by an extensive developer ecosystem.
Historique du momentum
19.05.17.08.
Fonctionnalités
| Image Resolution (Max.) | Native training resolution and default output: 512×512 pixels; higher resolutions (e.g., 768×768) technically possible, but with significantly increased VRAM requirements. |