tutorials/image/hidream/hidream-e1

[giscus[bot]]

tutorials/image/hidream/hidream-e1

This guide will help you understand and complete the ComfyUI native HiDream-I1 text-to-image workflow example

https://docs.comfy.org/tutorials/image/hidream/hidream-e1?utm_source=substack&utm_medium=email

[KannManMachen]

HiDream E1.1 Workflow:
the node "CFGNorm" is missing and I can not install it via ComfyUI manager.

[KannManMachen]

UPDATE: now it works

I did not notice that my ComfyUI portable version did not update properly via the manager since almost 3 month! ...

Instead I had to update via opening CMD and run:

git restore main.py
git checkout master
git pull

(inside of my "...\ComfyUI_windows_portable\ComfyUI" folder)

[comfyui-wiki]

I think I have already added an update notice at the beginning of this tutorial.

[comfyui-wiki]

For updating ComfyUI, you can refer to this link: https://docs.comfy.org/installation/update_comfyui

[olafrv]

For those with low VRAM (e.g. RTX 3070 Ti 8GB)...

This tutorial (Minute >=19) with defaults works (but way too slow) :
https://www.youtube.com/watch?v=dHdjjwWzrjM

Use original text encoders and vae from:
https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/tree/main/split_files

Use ND911 Q2_K GGUF difussion model and example workflow:
https://huggingface.co/ND911/HiDream_E1_1_bf16_ggufs/tree/main

Relevant file locations:

• "C:\ComfyUI\models\diffusion_models\hidream_e1_1_bf16_Q2_K.gguf"
• "C:\ComfyUI\models\text_encoders\clip_l_hidream.safetensors"
• "C:\ComfyUI\models\text_encoders\clip_g_hidream.safetensors"
• "C:\ComfyUI\models\text_encoders\t5xxl_fp8_e4m3fn_scaled.safetensors"
• "C:\ComfyUI\models\text_encoders\llama_3.1_8b_instruct_fp8_scaled.safetensors"
• "C:\ComfyUI\models\vae\ae.safetensors"