File size: 1,187 Bytes
4ffaba5 b5dce7a 4ffaba5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 |
---
license: apache-2.0
language:
- zh
- en
base_model:
- Qwen/Qwen2.5-32B-Instruct
datasets:
- anthracite-org/kalo-opus-instruct-22k-no-refusal
- unalignment/toxic-dpo-v0.2
- Orion-zhen/dpo-toxic-zh
library_name: transformers
---

GGUF Q-4_K_M Quantization of AiCloser/Qwen2.5-32B-AGI. Fit for 24GB card.
Modelcard included so that you can send it straight to ollama with `ollama create Qwen2.5-32B-AGI_Q4_K_M -f Modelfile` then `ollama run Qwen2.5-32B-AGI_Q4_K_M`
From my testing, it's not uncensored, but it's basically a lvl 1 guardrail and you just need to use a generic jailbreak to step around it.
________________________________
AGI means **A**spirational **G**rand **I**llusion
First Qwen2.5 32B Finetune, to fix its **Hypercensuritis**
> Hyper means high, and censura means censor, the suffix "-itis" is used to denote inflammation of a particular part or organ of the body. |