Lonecats Qwen 2.1 Fast Workflow w/ upscalers and post processing

LC123 Nodes: https://github.com/lonecatone23/ComfyUI_LC123_nodes

LC Vision Nodes: https://github.com/lonecatone23/ComfyUI_LC_Vision_nodes

LC note, which is tied to my LC vision pack, will automatically translate to the following languages:

• English

• 中文

• 繁體中文

• Русский

• 日本語

• 한국어

• Français

• Español

• عربي

• Türkçe

• Português (BR)

• فارسی

• עברית

• Italiano

Instagram: https://www.instagram.com/synth.studio.models/

Buy me a☕ https://ko-fi.com/lonecatone

This represents many hours of work that I give away for free. If you enjoy it, Please make sure to 👍like and post to my workflows. That is how you support me!!! Also, feel free to ⚡tip 😉

V2.0

So now that I've figured the model out....

• Built in optimizer settings custom designed for your computer with recommendations

• Removed unnecessary baggage in the image area like remove background and resize as it didn't work.

• Fixed the issues that every other workflow had with the inferior T2I Images. It now has it's own dedicated encoder

• Minor OCD stuff

V1.0 Beta.

I ran about 20 workflows and none of them were fast (or good) , so I spent some time figuring it out:

What's Included;

• Options for diffusion and GGUF models

• Optimizers (that work 🤔)

  • Model attention backend

  • Qwen image 2.1 Cache (standard)

  • EasyCache

  • I tried a few other's like Sage, Flash, and Sol, but the results were either negligable or they didn't function

• 5 step Turbo LoRA options

• Do Image Edit and T2I in the same workflow

• LC vision prompt assist and enhancers

• Multi image options

• VOSR 2.0 Upscaler

• Post Processing

What I found out along the way:

• Native node for Image edit and T2I sucks. If this model picks up traction, I'll build better ones

  • There is a bleed across the image inputs that picks up latent data somehow. No clue why

  • Still figuring this model out. At one point it was giving me fish gills and deer antlers. It's kinda weird

• Turbo LoRA gives decent results, but you really don't save a lot of time

• Qwen 2.1 SUCKS at prompt adherence