Hugging Face's logo

mradermacher
/
Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF

Transformers
GGUF
programming
code generation
code
codeqwen
Mixture of Experts
coding
coder
qwen2
chat
qwen
qwen-coder
mixture of experts
4 experts
2 active experts
40k context
qwen3
finetune
qwen3_moe
creative
all use cases
roleplay
Merge
imatrix
conversational
Model card Files
xet
Community
mradermacher commited on
Commit
b064f1f
·
verified ·
1 Parent(s): c94a68c

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +110 -0
README.md CHANGED
@@ -1,3 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 2 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: hf -->
@@ -7,3 +53,67 @@
7
  <!-- ### quants_skip: -->
8
  <!-- ### skip_mmproj: -->
9
  weighted/imatrix quants of https://huggingface.co/DavidAU/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: DavidAU/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2
3
+ language:
4
+ - en
5
+ - fr
6
+ - zh
7
+ - de
8
+ library_name: transformers
9
+ license: apache-2.0
10
+ mradermacher:
11
+ readme_rev: 1
12
+ quantized_by: mradermacher
13
+ tags:
14
+ - programming
15
+ - code generation
16
+ - code
17
+ - codeqwen
18
+ - programming
19
+ - code generation
20
+ - code
21
+ - codeqwen
22
+ - moe
23
+ - coding
24
+ - coder
25
+ - qwen2
26
+ - chat
27
+ - qwen
28
+ - qwen-coder
29
+ - chat
30
+ - qwen
31
+ - qwen-coder
32
+ - moe
33
+ - mixture of experts
34
+ - 4 experts
35
+ - 2 active experts
36
+ - 40k context
37
+ - qwen3
38
+ - finetune
39
+ - qwen3_moe
40
+ - creative
41
+ - all use cases
42
+ - roleplay
43
+ - merge
44
+ ---
45
+ ## About
46
+
47
  <!-- ### quantize_version: 2 -->
48
  <!-- ### output_tensor_quantised: 1 -->
49
  <!-- ### convert_type: hf -->
 
53
  <!-- ### quants_skip: -->
54
  <!-- ### skip_mmproj: -->
55
  weighted/imatrix quants of https://huggingface.co/DavidAU/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2
56
+
57
+ <!-- provided-files -->
58
+
59
+ ***For a convenient overview and download list, visit our [model page for this model](https://hf.tst.eu/model#Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF).***
60
+
61
+ static quants are available at https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-GGUF
62
+ ## Usage
63
+
64
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
65
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
66
+ more details, including on how to concatenate multi-part files.
67
+
68
+ ## Provided Quants
69
+
70
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
71
+
72
+ | Link | Type | Size/GB | Notes |
73
+ |:-----|:-----|--------:|:------|
74
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.imatrix.gguf) | imatrix | 0.1 | imatrix file (for creating your own qwuants) |
75
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ1_S.gguf) | i1-IQ1_S | 0.5 | for the desperate |
76
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ1_M.gguf) | i1-IQ1_M | 0.6 | mostly desperate |
77
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 0.6 | |
78
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ2_XS.gguf) | i1-IQ2_XS | 0.6 | |
79
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ2_S.gguf) | i1-IQ2_S | 0.7 | |
80
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ2_M.gguf) | i1-IQ2_M | 0.7 | |
81
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q2_K_S.gguf) | i1-Q2_K_S | 0.7 | very low quality |
82
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q2_K.gguf) | i1-Q2_K | 0.7 | IQ3_XXS probably better |
83
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 0.7 | lower quality |
84
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ3_XS.gguf) | i1-IQ3_XS | 0.8 | |
85
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ3_S.gguf) | i1-IQ3_S | 0.8 | beats Q3_K* |
86
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q3_K_S.gguf) | i1-Q3_K_S | 0.8 | IQ3_XS probably better |
87
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ3_M.gguf) | i1-IQ3_M | 0.8 | |
88
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q3_K_M.gguf) | i1-Q3_K_M | 0.9 | IQ3_S probably better |
89
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q3_K_L.gguf) | i1-Q3_K_L | 0.9 | IQ3_M probably better |
90
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ4_XS.gguf) | i1-IQ4_XS | 1.0 | |
91
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-IQ4_NL.gguf) | i1-IQ4_NL | 1.0 | prefer IQ4_XS |
92
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q4_0.gguf) | i1-Q4_0 | 1.0 | fast, low quality |
93
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q4_K_S.gguf) | i1-Q4_K_S | 1.0 | optimal size/speed/quality |
94
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q4_K_M.gguf) | i1-Q4_K_M | 1.1 | fast, recommended |
95
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q4_1.gguf) | i1-Q4_1 | 1.1 | |
96
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q5_K_S.gguf) | i1-Q5_K_S | 1.2 | |
97
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q5_K_M.gguf) | i1-Q5_K_M | 1.2 | |
98
+ | [GGUF](https://huggingface.co/mradermacher/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2-i1-GGUF/resolve/main/Qwen3-MOE-4x0.6B-2.4B-Writing-Thunder-V1.2.i1-Q6_K.gguf) | i1-Q6_K | 1.4 | practically like static Q6_K |
99
+
100
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
101
+ types (lower is better):
102
+
103
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
104
+
105
+ And here are Artefact2's thoughts on the matter:
106
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
107
+
108
+ ## FAQ / Model Request
109
+
110
+ See https://huggingface.co/mradermacher/model_requests for some answers to
111
+ questions you might have and/or if you want some other model quantized.
112
+
113
+ ## Thanks
114
+
115
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
116
+ me use its servers and providing upgrades to my workstation to enable
117
+ 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.
118
+
119
+ <!-- end -->