mradermacher commited on
Commit
c7294dd
1 Parent(s): c40d899

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +98 -0
README.md CHANGED
@@ -1,6 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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/aya-expanse-32b-abliterated
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: huihui-ai/aya-expanse-32b-abliterated
3
+ extra_gated_fields:
4
+ Affiliation: text
5
+ Country: country
6
+ I agree to use this model for non-commercial use ONLY: checkbox
7
+ Name: text
8
+ extra_gated_prompt: By submitting this form, you agree to the [License Agreement](https://cohere.com/c4ai-cc-by-nc-license) and
9
+ acknowledge that the information you provide will be collected, used, and shared
10
+ in accordance with Cohere’s [Privacy Policy]( https://cohere.com/privacy). You’ll
11
+ receive email updates about C4AI and Cohere research, events, products and services.
12
+ You can unsubscribe at any time.
13
+ language:
14
+ - en
15
+ - fr
16
+ - de
17
+ - es
18
+ - it
19
+ - pt
20
+ - ja
21
+ - ko
22
+ - zh
23
+ - ar
24
+ - el
25
+ - fa
26
+ - pl
27
+ - id
28
+ - cs
29
+ - he
30
+ - hi
31
+ - nl
32
+ - ro
33
+ - ru
34
+ - tr
35
+ - uk
36
+ - vi
37
+ library_name: transformers
38
+ license: cc-by-nc-4.0
39
+ quantized_by: mradermacher
40
+ tags:
41
+ - abliterated
42
+ - uncensored
43
+ ---
44
+ ## About
45
+
46
  <!-- ### quantize_version: 2 -->
47
  <!-- ### output_tensor_quantised: 1 -->
48
  <!-- ### convert_type: hf -->
49
  <!-- ### vocab_type: -->
50
  <!-- ### tags: nicoboss -->
51
  weighted/imatrix quants of https://huggingface.co/huihui-ai/aya-expanse-32b-abliterated
52
+
53
+ <!-- provided-files -->
54
+ static quants are available at https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-GGUF
55
+ ## Usage
56
+
57
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
58
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
59
+ more details, including on how to concatenate multi-part files.
60
+
61
+ ## Provided Quants
62
+
63
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
64
+
65
+ | Link | Type | Size/GB | Notes |
66
+ |:-----|:-----|--------:|:------|
67
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ1_M.gguf) | i1-IQ1_M | 8.5 | mostly desperate |
68
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 9.5 | |
69
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ2_XS.gguf) | i1-IQ2_XS | 10.4 | |
70
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ2_S.gguf) | i1-IQ2_S | 10.9 | |
71
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ2_M.gguf) | i1-IQ2_M | 11.7 | |
72
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q2_K_S.gguf) | i1-Q2_K_S | 12.1 | very low quality |
73
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q2_K.gguf) | i1-Q2_K | 12.9 | IQ3_XXS probably better |
74
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 13.1 | lower quality |
75
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q3_K_S.gguf) | i1-Q3_K_S | 14.8 | IQ3_XS probably better |
76
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ3_M.gguf) | i1-IQ3_M | 15.3 | |
77
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q3_K_M.gguf) | i1-Q3_K_M | 16.3 | IQ3_S probably better |
78
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q3_K_L.gguf) | i1-Q3_K_L | 17.7 | IQ3_M probably better |
79
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-IQ4_XS.gguf) | i1-IQ4_XS | 17.9 | |
80
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q4_K_S.gguf) | i1-Q4_K_S | 18.9 | optimal size/speed/quality |
81
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q4_K_M.gguf) | i1-Q4_K_M | 19.9 | fast, recommended |
82
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q5_K_S.gguf) | i1-Q5_K_S | 22.6 | |
83
+ | [GGUF](https://huggingface.co/mradermacher/aya-expanse-32b-abliterated-i1-GGUF/resolve/main/aya-expanse-32b-abliterated.i1-Q6_K.gguf) | i1-Q6_K | 26.6 | practically like static Q6_K |
84
+
85
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
86
+ types (lower is better):
87
+
88
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
89
+
90
+ And here are Artefact2's thoughts on the matter:
91
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
92
+
93
+ ## FAQ / Model Request
94
+
95
+ See https://huggingface.co/mradermacher/model_requests for some answers to
96
+ questions you might have and/or if you want some other model quantized.
97
+
98
+ ## Thanks
99
+
100
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
101
+ me use its servers and providing upgrades to my workstation to enable
102
+ 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.
103
+
104
+ <!-- end -->