VRAM 6GB(GeForce 1660 Ti)ã§ã‚‚Stable Diffusionã—ãŸã„ - ＷＩＣã®ä¸ã‹ã‚‰å··ã§å¤§äººæ°—ã®ã‚ªãƒ¼ãƒ—ンソースãªç”»åƒç”ŸæˆAI「Stable Diffusionã€ã‚’試ã—ã¦ã¿ãŸã„。 ã—ã‹ã—READ MEã«ã¯"the model is relatively lightweight and runs on a GPU with at least 10GB VRAM"ã¨ã‚る。ウãƒã®PCãŒç©ã‚“ã§ã‚‹ã®ã¯GeForce 1660 Ti 6GB...10GBã®VRAMã¯AIç•Œã«ã¨ã£ã¦å¤§ã—ãŸã‚‚ã®ã§ã¯ãªã„ã®ã ã‚ã†ãŒã€ä¸€èˆ¬äººã‹ã‚‰ã™ã‚‹ã¨ã‚ªãƒ¼ãƒãƒ¼ã‚¹ãƒšãƒƒã‚¯ãªã®ã 。 ãŸã 調ã¹ã¦ã¿ã‚‹ã¨VRAMä¸è¶³ã¸ã®å¯¾å‡¦æ³•ãŒã‚ã‚Šã€ç„¡äº‹ã«Stable Diffusionã§ç”»åƒç”Ÿæˆã§ããŸã€‚ ãã‚“ãªã‚ã‘ã§å‚™å¿˜éŒ²ã‚’ã“ã¡ã‚‰ã«æ®‹ã—ã¦ãŠã。 Pythonã¨GPU仕様環境を整ãˆã‚‹ モデルã®æº–備をã™ã‚‹ トラブル対応 出力例 txt2img img2img ã¾ã¨ã‚ Pythonã¨GPU仕様環境を整ãˆã‚‹ 下記記事通りPyTorchã¨CUDA1
GitHub - CompVis/stable-diffusion: A latent text-to-image diffusion model
Stable Diffusionã‚’ã„らã™ã¨ã‚„ã§ãƒ•ã‚¡ã‚¤ãƒ³ãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°ã™ã‚‹
Stable Diffusionã§ãƒ•ã‚¡ã‚¤ãƒ³ãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°(Finetune)を試㙠- Sheltie Garage Tech
{{#tags}}- {{label}}
{{/tags}}