mradermacher commited on
Commit
919cf6a
·
verified ·
1 Parent(s): 5f9d566

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +60 -0
README.md CHANGED
@@ -1,6 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 2 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: hf -->
4
  <!-- ### vocab_type: -->
5
  <!-- ### tags: nicoboss -->
6
  static quants of https://huggingface.co/Blackroot/Mirai-70B-1.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Blackroot/Mirai-70B-1.0
3
+ language:
4
+ - en
5
+ library_name: transformers
6
+ quantized_by: mradermacher
7
+ tags:
8
+ - roleplay
9
+ - experimental
10
+ - indie
11
+ - merge
12
+ ---
13
+ ## About
14
+
15
  <!-- ### quantize_version: 2 -->
16
  <!-- ### output_tensor_quantised: 1 -->
17
  <!-- ### convert_type: hf -->
18
  <!-- ### vocab_type: -->
19
  <!-- ### tags: nicoboss -->
20
  static quants of https://huggingface.co/Blackroot/Mirai-70B-1.0
21
+
22
+ <!-- provided-files -->
23
+ weighted/imatrix quants are available at https://huggingface.co/mradermacher/Mirai-70B-1.0-i1-GGUF
24
+ ## Usage
25
+
26
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
27
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
28
+ more details, including on how to concatenate multi-part files.
29
+
30
+ ## Provided Quants
31
+
32
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
33
+
34
+ | Link | Type | Size/GB | Notes |
35
+ |:-----|:-----|--------:|:------|
36
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q2_K.gguf) | Q2_K | 26.5 | |
37
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q3_K_S.gguf) | Q3_K_S | 31.0 | |
38
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q3_K_M.gguf) | Q3_K_M | 34.4 | lower quality |
39
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q3_K_L.gguf) | Q3_K_L | 37.2 | |
40
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q4_K_S.gguf) | Q4_K_S | 40.4 | fast, recommended |
41
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q4_K_M.gguf) | Q4_K_M | 42.6 | fast, recommended |
42
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q5_K_S.gguf) | Q5_K_S | 48.8 | |
43
+ | [GGUF](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q5_K_M.gguf) | Q5_K_M | 50.0 | |
44
+ | [PART 1](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q6_K.gguf.part2of2) | Q6_K | 58.0 | very good quality |
45
+ | [PART 1](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q8_0.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mirai-70B-1.0-GGUF/resolve/main/Mirai-70B-1.0.Q8_0.gguf.part2of2) | Q8_0 | 75.1 | fast, best quality |
46
+
47
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
48
+ types (lower is better):
49
+
50
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
51
+
52
+ And here are Artefact2's thoughts on the matter:
53
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
54
+
55
+ ## FAQ / Model Request
56
+
57
+ See https://huggingface.co/mradermacher/model_requests for some answers to
58
+ questions you might have and/or if you want some other model quantized.
59
+
60
+ ## Thanks
61
+
62
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
63
+ me use its servers and providing upgrades to my workstation to enable
64
+ 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.
65
+
66
+ <!-- end -->