Moe

Moe

GLM 4.7 Flash: Speed & Results Summary

Core Benefits (In 30 Seconds) What You Need to Start (Quick Setup) Critical Results (Tested on RTX 6000 Blackwell) Metric Value Prompt Speed >2,000 tokens/sec Generation Speed ~97 tokens/sec VRAM Usage 18% lower than standard Key Warning Pro Tip Monitor…

Problems with Kohya_ss SDXL Lora Merging

As i currently merge Illustrious SDXL LORAs with Kohya_ss I came along two annoying errormessages stopping me from creating them. As this annoyed me and I like to share solutions here my findings. While LORA merging If you run into…

Ana changes in OW2 Stadium Season 19

I must say that Ana got brain surgery in this season. It is not really a nerf but more a redesign of her usability and combos. Bad Stuff The old Sleeper build where you could have like 1000 HP is…

Is Qwen3 30b good? My experience and learning curve

As a lot of AI Models pop up and go I took interest in QWEN developed by Alibaba Group’s Qwen Team, which is capable of processing and generating text, images, audio, and video. Most of the time I had mixed…

Fixing a Numpy Error in ComfyUI

You maybe have seen this error “Numpy not found” after you upgraded your ComfyUI Version or used a new Workflow. As this errormessage is annoying a short explanation what happened before we go to the fix. The packages that you…