Summary
Full Transcript
In this episode, we set up a clean, portable ComfyUI with the latest Nunchaku nodes, then build and test practical workflows for Qwen ControlNet, Qwen Edit, and inpainting. You’ll see a fresh install from scratch, how to map your existing model folders, and how to run it. Testing Flux workflows (text-to-image, image-to-image, LoRA stack), and finish with real-world ControlNet use cases (canny, depth, pose) plus Qwen Edit single-image, two-image stitch, and masked inpaint. What you’ll learn: - Reliable fresh install of a portable ComfyUI + Nunchaku (no leftover conflicts) - Linking existing model paths to avoid duplicates - Installing Nunchaku wheels & optional Sage Attention for speed - Flux T2I/I2I with upscaling, tiling tips, VRAM presets - Qwen for Nunchaku: model choices (INT4/FP4, 32 vs 128, lightning), seeds, sizing, quality tweaks - Qwen ControlNet: canny, depth, pose - Qwen ControlNet Inpaint - Qwen Edit: instruction-style prompts (replace/add/change), two-image edits, and inpaint with masks Get the workflows and instructions from discord for free Step 1: use this link to join if you didn't join already https://discord.gg/gggpkVgBf3 Step 2: go to this post to get the workflows https://discord.com/channels/1245221993746399232/1417492626067947520/1417493862473138246 Get the easy ComfyUI Installer from here https://github.com/Tavris1/ComfyUI-Easy-Install Check Other Episodes https://www.youtube.com/playlist?list=PL-pohOSaL8P9kLZP8tQ1K1QWdZEgwiBM0 Unlock exclusive perks by joining our channel: https://www.youtube.com/channel/UCmMbwA-s3GZDKVzGZ-kPwaQ/join Grab your Pixaroma merch from here: https://www.youtube.com/@pixaroma/store #comfyui #comfyuitutorial #aiimageediting
