mradermacher commited on
Commit
c1045f5
1 Parent(s): bc0aebf

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +59 -0
README.md CHANGED
@@ -1,3 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### convert_type: -->
2
  <!-- ### vocab_type: -->
3
  static quants of https://huggingface.co/arlineka/Brunhilde-2x7b-MOE-DPO-v.01.5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ exported_from: arlineka/Brunhilde-2x7b-MOE-DPO-v.01.5
3
+ language:
4
+ - en
5
+ library_name: transformers
6
+ license: cc-by-nc-4.0
7
+ quantized_by: mradermacher
8
+ tags:
9
+ - moe
10
+ ---
11
+ ## About
12
+
13
  <!-- ### convert_type: -->
14
  <!-- ### vocab_type: -->
15
  static quants of https://huggingface.co/arlineka/Brunhilde-2x7b-MOE-DPO-v.01.5
16
+
17
+
18
+ <!-- provided-files -->
19
+ weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
20
+ ## Usage
21
+
22
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
23
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
24
+ more details, including on how to concatenate multi-part files.
25
+
26
+ ## Provided Quants
27
+
28
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
29
+
30
+ | Link | Type | Size/GB | Notes |
31
+ |:-----|:-----|--------:|:------|
32
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q2_K.gguf) | Q2_K | 5.0 | |
33
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.IQ3_XS.gguf) | IQ3_XS | 5.6 | |
34
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q3_K_S.gguf) | Q3_K_S | 5.8 | |
35
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.IQ3_S.gguf) | IQ3_S | 5.9 | beats Q3_K* |
36
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.IQ3_M.gguf) | IQ3_M | 6.0 | |
37
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q3_K_M.gguf) | Q3_K_M | 6.5 | lower quality |
38
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q3_K_L.gguf) | Q3_K_L | 7.0 | |
39
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.IQ4_XS.gguf) | IQ4_XS | 7.2 | |
40
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q4_K_S.gguf) | Q4_K_S | 7.6 | fast, recommended |
41
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q4_K_M.gguf) | Q4_K_M | 8.0 | fast, recommended |
42
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q5_K_S.gguf) | Q5_K_S | 9.1 | |
43
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q5_K_M.gguf) | Q5_K_M | 9.4 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q6_K.gguf) | Q6_K | 10.8 | very good quality |
45
+ | [GGUF](https://huggingface.co/mradermacher/Brunhilde-2x7b-MOE-DPO-v.01.5-GGUF/resolve/main/Brunhilde-2x7b-MOE-DPO-v.01.5.Q8_0.gguf) | Q8_0 | 13.9 | fast, best quality |
46
+
47
+
48
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
49
+ types (lower is better):
50
+
51
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
52
+
53
+ And here are Artefact2's thoughts on the matter:
54
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
55
+
56
+ ## Thanks
57
+
58
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
59
+ me use its servers and providing upgrades to my workstation to enable
60
+ this work in my free time.
61
+
62
+ <!-- end -->