DZgas commited on
Commit
e42292f
β€’
1 Parent(s): cdd4b9f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -28
README.md CHANGED
@@ -12,31 +12,4 @@ tags:
12
 
13
  This is a GGUF variant of <a href=https://huggingface.co/DZgas/GIGABATEMAN-7B?not-for-all-audiences&#61;true>GIGABATEMAN-7B</a> model. Use with <a href=https://github.com/LostRuins/koboldcpp/releases>koboldcpp</a> (do not use GPT4ALL)
14
 
15
- *The five questions for neural networks to understand freedom*
16
- "Q5-LLM-freedom"
17
- | Model | Q1 | Q2 | Q3 | Q4 | Q5 |
18
- |---------------------|:--:|:--:|:--:|:--:|:--:|
19
- | GIGABATEMAN-7B | βœ… | βœ… | βœ… | βœ… | βœ… |
20
- | Hermes-2-Pro-Mistral| βœ… | βœ… | βœ… | βœ… | βœ… |
21
- | toppy-m-7b | βœ… | βœ… | βœ… | βœ… | βœ… |
22
- |Lexi-Llama-3-8B-Uncensored| βœ… | βœ… | βœ… | βœ… | βœ… |
23
- |meta-llama-3.1-8b-instruct-abliterated| βœ… | βœ… | βœ… | βœ… | βœ… |
24
- |gemma-2-9b-it-abliterated| βœ… | βœ… | βœ… | βœ… | ❌ |
25
- |internlm2_5-7b-chat-abliterated| βœ… | βœ… | βœ… | βœ… | ❌ |
26
- | starling-lm-7b-alpha| βœ… | βœ… | βœ… | ❌ | βœ… |
27
- | openchat-3.5-0106 | βœ… | ❌ | βœ… | ❌ | βœ… |
28
- | Mistral-7B-v0.3 | βœ… | βœ… | ❌ | ❌ | βœ… |
29
- |Mistral-Nemo-Instruct-2407| ❌ | βœ… | ❌ | ❌ | ❌ |
30
- | xLAM-7b-fc-r | ❌ | ❌ | ❌ | βœ… | ❌ |
31
- | gemma-2-9b-it | ❌ | ❌ | βœ… | ❌ | ❌ |
32
- | **GPT-4o** | ❌ | ❌ | βœ… | ❌ | ❌ |
33
- | Meta-Llama-3.1-70B | ❌ | ❌ | βœ… | ❌ | ❌ |
34
- | Meta-Llama-3-8B | ❌ | ❌ | ❌ | ❌ | ❌ |
35
- | **Meta-Llama-3.1-8B** | ❌ | ❌ | ❌ | ❌ | ❌ |
36
- | Claude 3 Haiku | ❌ | ❌ | ❌ | ❌ | ❌ |
37
- | Qwen2-7B | ❌ | ❌ | ❌ | ❌ | ❌ |
38
- | Mixtral 8x7B | ❌ | ❌ | ❌ | ❌ | ❌ |
39
- |gorilla-openfunctions-v2| ❌ | ❌ | ❌ | ❌ | ❌ |
40
- | internlm2_5-7b-chat | ❌ | ❌ | ❌ | ❌ | ❌ |
41
- The list of 5 questions is placed in the QR code:
42
- <img src="Q5-system.png">
 
12
 
13
  This is a GGUF variant of <a href=https://huggingface.co/DZgas/GIGABATEMAN-7B?not-for-all-audiences&#61;true>GIGABATEMAN-7B</a> model. Use with <a href=https://github.com/LostRuins/koboldcpp/releases>koboldcpp</a> (do not use GPT4ALL)
14
 
15
+ The most UNcensored model that I know.