Friday’s big release was Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM from Alibaba’s Qwen research lab. I’ve been looking forward to this one: 27B is an …
i dont want to share it. but it was easy to add a little more description for basic criteria in the request. although, i did not detail a step-by-step process for drawing.
i had this idea to feed it an actual image, to base the drawing on. but i wasnt able to. llama.cpp seems to be messed up[esp vulkan pkg](and also i probably dont know the proper way to utilize features(if they even exist for a model))
i dont want to share it. but it was easy to add a little more description for basic criteria in the request. although, i did not detail a step-by-step process for drawing.
i had this idea to feed it an actual image, to base the drawing on. but i wasnt able to. llama.cpp seems to be messed up[esp vulkan pkg](and also i probably dont know the proper way to utilize features(if they even exist for a model))
Here’s the result by Qwen3.8-27b Q8 for “Create an SVG of an Iguana.”:
For your image with “Create an SVG of an Iguana based on the provided image.”:
(Using llama.cpp with Opencode)
hell yeah