From 772de55ee5e2fd84b9c08186c0926caffe73b483 Mon Sep 17 00:00:00 2001 From: GiusTex <112352961+GiusTex@users.noreply.github.com> Date: Sun, 1 Dec 2024 01:43:54 +0100 Subject: [PATCH] change clip models links Thanks to zmwv823 suggestion: https://github.com/GiusTex/ComfyUI-DiffusersImageOutpaint/issues/22#issuecomment-2508858557 --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index bec81ca..8e0c40d 100644 --- a/README.md +++ b/README.md @@ -32,7 +32,7 @@ ComfyUI nodes for outpainting images with diffusers, based on [diffusers-image-o - `model_index.json` ([example](https://huggingface.co/SG161222/RealVisXL_V5.0_Lightning/blob/main/model_index.json)) - controlnet_name: - `config_promax.json` ([example](https://huggingface.co/xinsir/controlnet-union-sdxl-1.0/blob/main/config_promax.json)), `diffusion_pytorch_model_promax.safetensors` ([example](https://huggingface.co/xinsir/controlnet-union-sdxl-1.0/blob/main/diffusion_pytorch_model_promax.safetensors)) -- (Dual) Clip Loader node: if you use the Clip Loader instead of Checkpoint Loader Simple, and want to use RealVisXL_V5.0_Lightning, it works with [`clip_I`](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/text_encoder/model.fp16.safetensors) and [`model.fp16`](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/text_encoder_2/model.fp16.safetensors) (from sdxl-base), and `sdxl type`; you can use [this workflow](https://github.com/GiusTex/ComfyUI-DiffusersImageOutpaint/blob/New-Pad-Node-Options/Diffusers-Outpaint-DoubleWorkflow.json). +- (Dual) Clip Loader node: if you use the Clip Loader instead of Checkpoint Loader Simple, and want to use an `sdxl type` model like RealVisXL_V5.0_Lightning, you can download `clip_I` and `clip_g` from [here](https://huggingface.co/Comfy-Org/stable-diffusion-3.5-fp8/tree/main/text_encoders). You can use [this workflow](https://github.com/GiusTex/ComfyUI-DiffusersImageOutpaint/blob/New-Pad-Node-Options/Diffusers-Outpaint-DoubleWorkflow.json) (change model.fp16 with `clip_g`). ## Overview - **Minimum VRAM**: 6 gb with 1280x720 image, rtx 3060, RealVisXL_V5.0_Lightning, sdxl-vae-fp16-fix, controlnet-union-sdxl-promax using `sequential_cpu_offload`, otherwise 8,3 gb;