Hugging Face's logo

mradermacher
/
Llama3.2-3B-ShiningValiant2-i1-GGUF

Transformers
GGUF
English
shining-valiant
shining-valiant-2
valiant
valiant-labs
llama
llama-3.2
llama-3.2-instruct
llama-3.2-instruct-3b
llama-3
llama-3-instruct
llama-3-instruct-3b
3b
science
physics
biology
chemistry
compsci
computer-science
engineering
technical
conversational
chat
instruct
Inference Endpoints
imatrix
Model card Files Community
mradermacher commited on
Commit
5d33e6e
•
1 Parent(s): 4f25bc7

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +99 -0
README.md CHANGED
@@ -1,6 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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/ValiantLabs/Llama3.2-3B-ShiningValiant2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: ValiantLabs/Llama3.2-3B-ShiningValiant2
3
+ datasets:
4
+ - sequelbox/Celestia
5
+ - sequelbox/Spurline
6
+ - sequelbox/Supernova
7
+ language:
8
+ - en
9
+ library_name: transformers
10
+ license: llama3.2
11
+ model_type: llama
12
+ quantized_by: mradermacher
13
+ tags:
14
+ - shining-valiant
15
+ - shining-valiant-2
16
+ - valiant
17
+ - valiant-labs
18
+ - llama
19
+ - llama-3.2
20
+ - llama-3.2-instruct
21
+ - llama-3.2-instruct-3b
22
+ - llama-3
23
+ - llama-3-instruct
24
+ - llama-3-instruct-3b
25
+ - 3b
26
+ - science
27
+ - physics
28
+ - biology
29
+ - chemistry
30
+ - compsci
31
+ - computer-science
32
+ - engineering
33
+ - technical
34
+ - conversational
35
+ - chat
36
+ - instruct
37
+ ---
38
+ ## About
39
+
40
  <!-- ### quantize_version: 2 -->
41
  <!-- ### output_tensor_quantised: 1 -->
42
  <!-- ### convert_type: hf -->
43
  <!-- ### vocab_type: -->
44
  <!-- ### tags: nicoboss -->
45
  weighted/imatrix quants of https://huggingface.co/ValiantLabs/Llama3.2-3B-ShiningValiant2
46
+
47
+ <!-- provided-files -->
48
+ static quants are available at https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-GGUF
49
+ ## Usage
50
+
51
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
52
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
53
+ more details, including on how to concatenate multi-part files.
54
+
55
+ ## Provided Quants
56
+
57
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
58
+
59
+ | Link | Type | Size/GB | Notes |
60
+ |:-----|:-----|--------:|:------|
61
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ1_S.gguf) | i1-IQ1_S | 1.0 | for the desperate |
62
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ1_M.gguf) | i1-IQ1_M | 1.0 | mostly desperate |
63
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 1.1 | |
64
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ2_XS.gguf) | i1-IQ2_XS | 1.2 | |
65
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ2_S.gguf) | i1-IQ2_S | 1.3 | |
66
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ2_M.gguf) | i1-IQ2_M | 1.3 | |
67
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 1.4 | lower quality |
68
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q2_K.gguf) | i1-Q2_K | 1.5 | IQ3_XXS probably better |
69
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ3_XS.gguf) | i1-IQ3_XS | 1.6 | |
70
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ3_S.gguf) | i1-IQ3_S | 1.6 | beats Q3_K* |
71
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q3_K_S.gguf) | i1-Q3_K_S | 1.6 | IQ3_XS probably better |
72
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ3_M.gguf) | i1-IQ3_M | 1.7 | |
73
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q3_K_M.gguf) | i1-Q3_K_M | 1.8 | IQ3_S probably better |
74
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q3_K_L.gguf) | i1-Q3_K_L | 1.9 | IQ3_M probably better |
75
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-IQ4_XS.gguf) | i1-IQ4_XS | 1.9 | |
76
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q4_0_4_4.gguf) | i1-Q4_0_4_4 | 2.0 | fast on arm, low quality |
77
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q4_0_4_8.gguf) | i1-Q4_0_4_8 | 2.0 | fast on arm+i8mm, low quality |
78
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q4_0_8_8.gguf) | i1-Q4_0_8_8 | 2.0 | fast on arm+sve, low quality |
79
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q4_0.gguf) | i1-Q4_0 | 2.0 | fast, low quality |
80
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q4_K_S.gguf) | i1-Q4_K_S | 2.0 | optimal size/speed/quality |
81
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q4_K_M.gguf) | i1-Q4_K_M | 2.1 | fast, recommended |
82
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q5_K_S.gguf) | i1-Q5_K_S | 2.4 | |
83
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q5_K_M.gguf) | i1-Q5_K_M | 2.4 | |
84
+ | [GGUF](https://huggingface.co/mradermacher/Llama3.2-3B-ShiningValiant2-i1-GGUF/resolve/main/Llama3.2-3B-ShiningValiant2.i1-Q6_K.gguf) | i1-Q6_K | 2.7 | practically like static Q6_K |
85
+
86
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
87
+ types (lower is better):
88
+
89
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
90
+
91
+ And here are Artefact2's thoughts on the matter:
92
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
93
+
94
+ ## FAQ / Model Request
95
+
96
+ See https://huggingface.co/mradermacher/model_requests for some answers to
97
+ questions you might have and/or if you want some other model quantized.
98
+
99
+ ## Thanks
100
+
101
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
102
+ me use its servers and providing upgrades to my workstation to enable
103
+ 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.
104
+
105
+ <!-- end -->