mradermacher commited on
Commit
c3f758e
1 Parent(s): f260660

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +53 -0
README.md CHANGED
@@ -1 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  static quants of https://huggingface.co/perlthoughts/Starling-LM-alpha-8x7B-MoE
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ datasets:
3
+ - berkeley-nest/Nectar
4
+ language:
5
+ - en
6
+ library_name: transformers
7
+ license: cc-by-nc-4.0
8
+ quantized_by: mradermacher
9
+ tags:
10
+ - reward model
11
+ - RLHF
12
+ - RLAIF
13
+ - moe
14
+ ---
15
+ ## About
16
+
17
  static quants of https://huggingface.co/perlthoughts/Starling-LM-alpha-8x7B-MoE
18
+
19
+
20
+ <!-- provided-files -->
21
+ ## Usage
22
+
23
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
24
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
25
+ more details, including on how to concatenate multi-part files.
26
+
27
+ ## Provided Quants
28
+
29
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
30
+
31
+ | Link | Type | Size/GB | Notes |
32
+ |:-----|:-----|--------:|:------|
33
+ | [GGUF](https://huggingface.co/mradermacher/Starling-LM-alpha-8x7B-MoE-GGUF/resolve/main/Starling-LM-alpha-8x7B-MoE.Q2_K.gguf) | Q2_K | 17.6 | |
34
+ | [GGUF](https://huggingface.co/mradermacher/Starling-LM-alpha-8x7B-MoE-GGUF/resolve/main/Starling-LM-alpha-8x7B-MoE.Q3_K_S.gguf) | Q3_K_S | 20.7 | |
35
+ | [GGUF](https://huggingface.co/mradermacher/Starling-LM-alpha-8x7B-MoE-GGUF/resolve/main/Starling-LM-alpha-8x7B-MoE.Q3_K_M.gguf) | Q3_K_M | 22.8 | lower quality |
36
+ | [GGUF](https://huggingface.co/mradermacher/Starling-LM-alpha-8x7B-MoE-GGUF/resolve/main/Starling-LM-alpha-8x7B-MoE.Q4_K_S.gguf) | Q4_K_S | 27.0 | fast, medium quality |
37
+ | [PART 1](https://huggingface.co/mradermacher/Starling-LM-alpha-8x7B-MoE-GGUF/resolve/main/Starling-LM-alpha-8x7B-MoE.Q8_0.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Starling-LM-alpha-8x7B-MoE-GGUF/resolve/main/Starling-LM-alpha-8x7B-MoE.Q8_0.gguf.part2of2) | Q8_0 | 49.8 | fast, best quality |
38
+
39
+
40
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
41
+ types (lower is better):
42
+
43
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
44
+
45
+ And here are Artefact2's thoughts on the matter:
46
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
47
+
48
+ ## Thanks
49
+
50
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
51
+ me use its servers and providing upgrades to my workstation to enable
52
+ this work in my free time.
53
+
54
+ <!-- end -->