mradermacher

/

Qwen_WhatTheFuck-GGUF

like 0

...but its from my own "stupidity-ai" organization. im confused now

What's wrong with a model getting released by "stupidity-ai"? mergekit-community has no monopoly on releasing merges. The mergekit-community doesn't have a model named "Qwen_WhatTheFuck". Can you please link what model you are referring to? We could to it as well if you don't like this one.

In the end Qwen_WhatTheFuck is just the following merge and who uploaded it really doesn't matter:

base_model: Qwen/Qwen2.5-0.5B-Instruct
dtype: bfloat16
merge_method: task_arithmetic
parameters:
  normalize: true
slices:
- sources:
  - layer_range: [0, 24]
    model: Qwen/Qwen2.5-0.5B-Instruct
  - layer_range: [0, 24]
    model: Qwen/Qwen2.5-0.5B-Instruct
    parameters:
      weight: [1.0, -1.0, 1.0, -1.0, 1.0]
  - layer_range: [0, 24]
    model: Qwen/Qwen2-0.5B-Instruct
    parameters:
      weight: [-1.0, 1.0, -1.0, 1.0, -1.0]

I see you are the only member of "stupidity-ai" so you must have uploaded it yourself. I guess you already have it so no need to download it but others might be interested in it. In fact, this quant has already accumulated 122 downloads in the past 12 hours.

it is a stupid merge, so stupidity-ai was made to host stupid merges.

the merge is surprisingly usable though.

i was just confused on how you even FOUND the model

@Austinkeith2010 clicking on "Models" at the top of the page gives you a handy model search listing new models :) (that's how we found models a yera ago, nowadays we have our own filter on essentially on top of that).

true