Hugging Face's logo

mradermacher
/
TowerInstruct-WMT24-Chat-7B-i1-GGUF

Transformers
GGUF
Inference Endpoints
imatrix
conversational
Model card Files Community
mradermacher commited on
Commit
513eb86
•
1 Parent(s): bcc8723

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +79 -0
README.md CHANGED
@@ -1,6 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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/Unbabel/TowerInstruct-WMT24-Chat-7B
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Unbabel/TowerInstruct-WMT24-Chat-7B
3
+ language:
4
+ - en
5
+ - de
6
+ - fr
7
+ - zh
8
+ - pt
9
+ - nl
10
+ - ru
11
+ - ko
12
+ - it
13
+ - es
14
+ library_name: transformers
15
+ license: cc-by-nc-4.0
16
+ quantized_by: mradermacher
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/Unbabel/TowerInstruct-WMT24-Chat-7B
26
+
27
+ <!-- provided-files -->
28
+ static quants are available at https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-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/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ1_S.gguf) | i1-IQ1_S | 1.6 | for the desperate |
42
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ1_M.gguf) | i1-IQ1_M | 1.8 | mostly desperate |
43
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 2.0 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ2_XS.gguf) | i1-IQ2_XS | 2.1 | |
45
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ2_S.gguf) | i1-IQ2_S | 2.3 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ2_M.gguf) | i1-IQ2_M | 2.5 | |
47
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q2_K.gguf) | i1-Q2_K | 2.6 | IQ3_XXS probably better |
48
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 2.7 | lower quality |
49
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ3_XS.gguf) | i1-IQ3_XS | 2.9 | |
50
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ3_S.gguf) | i1-IQ3_S | 3.0 | beats Q3_K* |
51
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q3_K_S.gguf) | i1-Q3_K_S | 3.0 | IQ3_XS probably better |
52
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ3_M.gguf) | i1-IQ3_M | 3.2 | |
53
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q3_K_M.gguf) | i1-Q3_K_M | 3.4 | IQ3_S probably better |
54
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q3_K_L.gguf) | i1-Q3_K_L | 3.7 | IQ3_M probably better |
55
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-IQ4_XS.gguf) | i1-IQ4_XS | 3.7 | |
56
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q4_0_4_4.gguf) | i1-Q4_0_4_4 | 3.9 | fast on arm, low quality |
57
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q4_0_4_8.gguf) | i1-Q4_0_4_8 | 3.9 | fast on arm+i8mm, low quality |
58
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q4_0_8_8.gguf) | i1-Q4_0_8_8 | 3.9 | fast on arm+sve, low quality |
59
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q4_0.gguf) | i1-Q4_0 | 3.9 | fast, low quality |
60
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q4_K_S.gguf) | i1-Q4_K_S | 4.0 | optimal size/speed/quality |
61
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q4_K_M.gguf) | i1-Q4_K_M | 4.2 | fast, recommended |
62
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q5_K_S.gguf) | i1-Q5_K_S | 4.8 | |
63
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q5_K_M.gguf) | i1-Q5_K_M | 4.9 | |
64
+ | [GGUF](https://huggingface.co/mradermacher/TowerInstruct-WMT24-Chat-7B-i1-GGUF/resolve/main/TowerInstruct-WMT24-Chat-7B.i1-Q6_K.gguf) | i1-Q6_K | 5.6 | practically like static Q6_K |
65
+
66
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
67
+ types (lower is better):
68
+
69
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
70
+
71
+ And here are Artefact2's thoughts on the matter:
72
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
73
+
74
+ ## FAQ / Model Request
75
+
76
+ See https://huggingface.co/mradermacher/model_requests for some answers to
77
+ questions you might have and/or if you want some other model quantized.
78
+
79
+ ## Thanks
80
+
81
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
82
+ me use its servers and providing upgrades to my workstation to enable
83
+ 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.
84
+
85
+ <!-- end -->