mradermacher commited on
Commit
5ace1dc
1 Parent(s): 28ca25d

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +76 -0
README.md CHANGED
@@ -1,6 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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/UsernameJustAnother/Nemo-12B-Marlin-v4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: UsernameJustAnother/Nemo-12B-Marlin-v4
3
+ language:
4
+ - en
5
+ library_name: transformers
6
+ license: apache-2.0
7
+ quantized_by: mradermacher
8
+ tags:
9
+ - text-generation-inference
10
+ - transformers
11
+ - unsloth
12
+ - mistral
13
+ - trl
14
+ - rp
15
+ - experimental
16
+ - long-context
17
+ ---
18
+ ## About
19
+
20
  <!-- ### quantize_version: 2 -->
21
  <!-- ### output_tensor_quantised: 1 -->
22
  <!-- ### convert_type: hf -->
23
  <!-- ### vocab_type: -->
24
  <!-- ### tags: nicoboss -->
25
  weighted/imatrix quants of https://huggingface.co/UsernameJustAnother/Nemo-12B-Marlin-v4
26
+
27
+ <!-- provided-files -->
28
+ static quants are available at https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-GGUF
29
+ ## Usage
30
+
31
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
32
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
33
+ more details, including on how to concatenate multi-part files.
34
+
35
+ ## Provided Quants
36
+
37
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
38
+
39
+ | Link | Type | Size/GB | Notes |
40
+ |:-----|:-----|--------:|:------|
41
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ1_S.gguf) | i1-IQ1_S | 3.1 | for the desperate |
42
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ1_M.gguf) | i1-IQ1_M | 3.3 | mostly desperate |
43
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 3.7 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ2_XS.gguf) | i1-IQ2_XS | 4.0 | |
45
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ2_S.gguf) | i1-IQ2_S | 4.2 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ2_M.gguf) | i1-IQ2_M | 4.5 | |
47
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q2_K.gguf) | i1-Q2_K | 4.9 | IQ3_XXS probably better |
48
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 5.0 | lower quality |
49
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ3_XS.gguf) | i1-IQ3_XS | 5.4 | |
50
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q3_K_S.gguf) | i1-Q3_K_S | 5.6 | IQ3_XS probably better |
51
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ3_S.gguf) | i1-IQ3_S | 5.7 | beats Q3_K* |
52
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ3_M.gguf) | i1-IQ3_M | 5.8 | |
53
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q3_K_M.gguf) | i1-Q3_K_M | 6.2 | IQ3_S probably better |
54
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q3_K_L.gguf) | i1-Q3_K_L | 6.7 | IQ3_M probably better |
55
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-IQ4_XS.gguf) | i1-IQ4_XS | 6.8 | |
56
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q4_0.gguf) | i1-Q4_0 | 7.2 | fast, low quality |
57
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q4_K_S.gguf) | i1-Q4_K_S | 7.2 | optimal size/speed/quality |
58
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q4_K_M.gguf) | i1-Q4_K_M | 7.6 | fast, recommended |
59
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q5_K_S.gguf) | i1-Q5_K_S | 8.6 | |
60
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q5_K_M.gguf) | i1-Q5_K_M | 8.8 | |
61
+ | [GGUF](https://huggingface.co/mradermacher/Nemo-12B-Marlin-v4-i1-GGUF/resolve/main/Nemo-12B-Marlin-v4.i1-Q6_K.gguf) | i1-Q6_K | 10.2 | practically like static Q6_K |
62
+
63
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
64
+ types (lower is better):
65
+
66
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
67
+
68
+ And here are Artefact2's thoughts on the matter:
69
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
70
+
71
+ ## FAQ / Model Request
72
+
73
+ See https://huggingface.co/mradermacher/model_requests for some answers to
74
+ questions you might have and/or if you want some other model quantized.
75
+
76
+ ## Thanks
77
+
78
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
79
+ me use its servers and providing upgrades to my workstation to enable
80
+ 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.
81
+
82
+ <!-- end -->