mradermacher commited on
Commit
80f8450
1 Parent(s): 88a536e

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +70 -0
README.md CHANGED
@@ -1,6 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
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/anthracite-org/magnum-v4-123b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: anthracite-org/magnum-v4-123b
3
+ language:
4
+ - en
5
+ library_name: transformers
6
+ license: other
7
+ license_name: mrl
8
+ quantized_by: mradermacher
9
+ tags:
10
+ - chat
11
+ ---
12
+ ## About
13
+
14
  <!-- ### quantize_version: 2 -->
15
  <!-- ### output_tensor_quantised: 1 -->
16
  <!-- ### convert_type: hf -->
17
  <!-- ### vocab_type: -->
18
  <!-- ### tags: nicoboss -->
19
  weighted/imatrix quants of https://huggingface.co/anthracite-org/magnum-v4-123b
20
+
21
+ <!-- provided-files -->
22
+ static quants are available at https://huggingface.co/mradermacher/magnum-v4-123b-GGUF
23
+ ## Usage
24
+
25
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
26
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
27
+ more details, including on how to concatenate multi-part files.
28
+
29
+ ## Provided Quants
30
+
31
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
32
+
33
+ | Link | Type | Size/GB | Notes |
34
+ |:-----|:-----|--------:|:------|
35
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ1_S.gguf) | i1-IQ1_S | 26.1 | for the desperate |
36
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ1_M.gguf) | i1-IQ1_M | 28.5 | mostly desperate |
37
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 32.5 | |
38
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ2_XS.gguf) | i1-IQ2_XS | 36.2 | |
39
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ2_S.gguf) | i1-IQ2_S | 38.5 | |
40
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ2_M.gguf) | i1-IQ2_M | 41.7 | |
41
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q2_K.gguf) | i1-Q2_K | 45.3 | IQ3_XXS probably better |
42
+ | [GGUF](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 47.1 | lower quality |
43
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ3_XS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ3_XS.gguf.part2of2) | i1-IQ3_XS | 50.2 | |
44
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q3_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q3_K_S.gguf.part2of2) | i1-Q3_K_S | 52.9 | IQ3_XS probably better |
45
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ3_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ3_S.gguf.part2of2) | i1-IQ3_S | 53.1 | beats Q3_K* |
46
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ3_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ3_M.gguf.part2of2) | i1-IQ3_M | 55.4 | |
47
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q3_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q3_K_M.gguf.part2of2) | i1-Q3_K_M | 59.2 | IQ3_S probably better |
48
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q3_K_L.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q3_K_L.gguf.part2of2) | i1-Q3_K_L | 64.7 | IQ3_M probably better |
49
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ4_XS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-IQ4_XS.gguf.part2of2) | i1-IQ4_XS | 65.5 | |
50
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q4_0.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q4_0.gguf.part2of2) | i1-Q4_0 | 69.4 | fast, low quality |
51
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q4_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q4_K_S.gguf.part2of2) | i1-Q4_K_S | 69.7 | optimal size/speed/quality |
52
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q4_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q4_K_M.gguf.part2of2) | i1-Q4_K_M | 73.3 | fast, recommended |
53
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q5_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q5_K_S.gguf.part2of2) | i1-Q5_K_S | 84.5 | |
54
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q5_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q5_K_M.gguf.part2of2) | i1-Q5_K_M | 86.6 | |
55
+ | [PART 1](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q6_K.gguf.part1of3) [PART 2](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q6_K.gguf.part2of3) [PART 3](https://huggingface.co/mradermacher/magnum-v4-123b-i1-GGUF/resolve/main/magnum-v4-123b.i1-Q6_K.gguf.part3of3) | i1-Q6_K | 100.7 | practically like static Q6_K |
56
+
57
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
58
+ types (lower is better):
59
+
60
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
61
+
62
+ And here are Artefact2's thoughts on the matter:
63
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
64
+
65
+ ## FAQ / Model Request
66
+
67
+ See https://huggingface.co/mradermacher/model_requests for some answers to
68
+ questions you might have and/or if you want some other model quantized.
69
+
70
+ ## Thanks
71
+
72
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
73
+ me use its servers and providing upgrades to my workstation to enable
74
+ 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.
75
+
76
+ <!-- end -->