Hugging Face's logo

mradermacher
/
EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF

Transformers
GGUF
English
Korean
lg-ai
exaone
exaone-3.5
abliterated
uncensored
Inference Endpoints
imatrix
conversational
Model card Files Community
mradermacher commited on
Commit
babcb39
•
1 Parent(s): beab477

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +77 -0
README.md CHANGED
@@ -1,6 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 2 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: hf -->
4
  <!-- ### vocab_type: -->
5
  <!-- ### tags: nicoboss -->
6
  weighted/imatrix quants of https://huggingface.co/huihui-ai/EXAONE-3.5-32B-Instruct-abliterated
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: huihui-ai/EXAONE-3.5-32B-Instruct-abliterated
3
+ language:
4
+ - en
5
+ - ko
6
+ library_name: transformers
7
+ license: other
8
+ license_link: LICENSE
9
+ license_name: exaone
10
+ quantized_by: mradermacher
11
+ tags:
12
+ - lg-ai
13
+ - exaone
14
+ - exaone-3.5
15
+ - abliterated
16
+ - uncensored
17
+ ---
18
+ ## About
19
+
20
  <!-- ### quantize_version: 2 -->
21
  <!-- ### output_tensor_quantised: 1 -->
22
  <!-- ### convert_type: hf -->
23
  <!-- ### vocab_type: -->
24
  <!-- ### tags: nicoboss -->
25
  weighted/imatrix quants of https://huggingface.co/huihui-ai/EXAONE-3.5-32B-Instruct-abliterated
26
+
27
+ <!-- provided-files -->
28
+ static quants are available at https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-GGUF
29
+ ## Usage
30
+
31
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
32
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
33
+ more details, including on how to concatenate multi-part files.
34
+
35
+ ## Provided Quants
36
+
37
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
38
+
39
+ | Link | Type | Size/GB | Notes |
40
+ |:-----|:-----|--------:|:------|
41
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ1_S.gguf) | i1-IQ1_S | 7.1 | for the desperate |
42
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ1_M.gguf) | i1-IQ1_M | 7.7 | mostly desperate |
43
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 8.8 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ2_XS.gguf) | i1-IQ2_XS | 9.7 | |
45
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ2_S.gguf) | i1-IQ2_S | 10.1 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ2_M.gguf) | i1-IQ2_M | 11.0 | |
47
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q2_K_S.gguf) | i1-Q2_K_S | 11.2 | very low quality |
48
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q2_K.gguf) | i1-Q2_K | 12.0 | IQ3_XXS probably better |
49
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 12.6 | lower quality |
50
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ3_XS.gguf) | i1-IQ3_XS | 13.4 | |
51
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q3_K_S.gguf) | i1-Q3_K_S | 14.1 | IQ3_XS probably better |
52
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ3_S.gguf) | i1-IQ3_S | 14.1 | beats Q3_K* |
53
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ3_M.gguf) | i1-IQ3_M | 14.5 | |
54
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q3_K_M.gguf) | i1-Q3_K_M | 15.6 | IQ3_S probably better |
55
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q3_K_L.gguf) | i1-Q3_K_L | 16.9 | IQ3_M probably better |
56
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-IQ4_XS.gguf) | i1-IQ4_XS | 17.3 | |
57
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q4_0.gguf) | i1-Q4_0 | 18.3 | fast, low quality |
58
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q4_K_S.gguf) | i1-Q4_K_S | 18.4 | optimal size/speed/quality |
59
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q4_K_M.gguf) | i1-Q4_K_M | 19.4 | fast, recommended |
60
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q5_K_S.gguf) | i1-Q5_K_S | 22.2 | |
61
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q5_K_M.gguf) | i1-Q5_K_M | 22.8 | |
62
+ | [GGUF](https://huggingface.co/mradermacher/EXAONE-3.5-32B-Instruct-abliterated-i1-GGUF/resolve/main/EXAONE-3.5-32B-Instruct-abliterated.i1-Q6_K.gguf) | i1-Q6_K | 26.4 | practically like static Q6_K |
63
+
64
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
65
+ types (lower is better):
66
+
67
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
68
+
69
+ And here are Artefact2's thoughts on the matter:
70
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
71
+
72
+ ## FAQ / Model Request
73
+
74
+ See https://huggingface.co/mradermacher/model_requests for some answers to
75
+ questions you might have and/or if you want some other model quantized.
76
+
77
+ ## Thanks
78
+
79
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
80
+ me use its servers and providing upgrades to my workstation to enable
81
+ this work in my free time. Additional thanks to [@nicoboss](https://huggingface.co/nicoboss) for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.
82
+
83
+ <!-- end -->