Transformers
GGUF
Chinese
English
Inference Endpoints
conversational
Daemonrat commited on
Commit
4ffaba5
1 Parent(s): d751eba

Update README.md

Browse files

![DALL·E 2024-11-15 11.31.45 - A hulking space marine in battle-torn armor, holding a rusty chainsaw pointed directly at the viewer in an action pose. The scene is depicted in retro.webp](https://cdn-uploads.huggingface.co/production/uploads/630450da0907b9a115c85ca7/mSDiAJOoP17UrL1bsqCsm.webp)

Files changed (1) hide show
  1. README.md +25 -3
README.md CHANGED
@@ -1,3 +1,25 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ language:
4
+ - zh
5
+ - en
6
+ base_model:
7
+ - Qwen/Qwen2.5-32B-Instruct
8
+ datasets:
9
+ - anthracite-org/kalo-opus-instruct-22k-no-refusal
10
+ - unalignment/toxic-dpo-v0.2
11
+ - Orion-zhen/dpo-toxic-zh
12
+ library_name: transformers
13
+ ---
14
+
15
+ ![DALL·E 2024-11-15 11.31.45 - A hulking space marine in battle-torn armor, holding a rusty chainsaw pointed directly at the viewer in an action pose. The scene is depicted in retro.webp](https://cdn-uploads.huggingface.co/production/uploads/630450da0907b9a115c85ca7/mSDiAJOoP17UrL1bsqCsm.webp)
16
+
17
+ 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`
18
+ 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.
19
+
20
+ ________________________________
21
+ AGI means **A**spirational **G**rand **I**llusion
22
+
23
+ First Qwen2.5 32B Finetune, to fix its **Hypercensuritis**
24
+
25
+ > Hyper means high, and censura means censor, the suffix "-itis" is used to denote inflammation of a particular part or organ of the body.