https://accordingly.dev/en/tags/qwen/ Recent content in Qwen on accordingly Hugo en-us Sun, 27 Sep 2026 09:00:00 +0200 https://accordingly.dev/en/posts/190-token-al-secondo-qwen36-35b-a3b-rtx5080/ Sun, 27 Sep 2026 09:00:00 +0200 https://accordingly.dev/en/posts/190-token-al-secondo-qwen36-35b-a3b-rtx5080/ A 14 GB quantized Qwen3.6-35B-A3B generates 190 tokens/s on a 16 GB RTX 5080 Laptop. MTP, q8_0 KV cache and the full numbers from the test campaign. https://accordingly.dev/en/posts/homelab-week-12-18-september-2026/ Fri, 18 Sep 2026 15:56:00 +0200 https://accordingly.dev/en/posts/homelab-week-12-18-september-2026/ The week in review: from the .NET/Angular CMS for Hugo to covers generated with Qwen Image, through the GX10-to-Strix Scar 16 link, down to the activity log that turns the work diary into post drafts. https://accordingly.dev/en/posts/automatic-covers-qwen-image-cms/ Fri, 18 Sep 2026 12:37:01 +0000 https://accordingly.dev/en/posts/automatic-covers-qwen-image-cms/ How I added automatic cover generation to my blog CMS: ComfyUI with Qwen Image 2512 on the GX10, an async endpoint with polling, and a RAM constraint that forced a design choice. https://accordingly.dev/en/posts/connecting-gx10-and-strix-scar-16-multimedia-tests-and-triggers/ Fri, 18 Sep 2026 10:49:55 +0200 https://accordingly.dev/en/posts/connecting-gx10-and-strix-scar-16-multimedia-tests-and-triggers/ Configuring Copilot on the Strix Scar 16 to talk to Qwen on the GX10, image-submission tests, and verifying the automated triggers. https://accordingly.dev/en/posts/qwen38-flash-next-on-gx10/ Thu, 17 Sep 2026 15:20:00 +0200 https://accordingly.dev/en/posts/qwen38-flash-next-on-gx10/ How I installed Qwen3.8 Flash Next with llama.cpp on the ASUS Ascent GX10, exposed it as a service with systemd+linger, and replaced Qwen3.6.