turboderp commited on
Commit
45df346
·
verified ·
1 Parent(s): 3184a1c

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +86 -0
README.md ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: Qwen/Qwen3-VL-8B-Instruct
4
+ base_model_relation: quantized
5
+ quantized_by: turboderp
6
+ tags:
7
+ - exl3
8
+ ---
9
+
10
+ EXL3 quants of [Qwen3-VL-32B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-32B-Instruct)
11
+
12
+ ⚠️ Requires ExLlamaV3 v0.0.13 (or v0.0.12 `dev` branch)
13
+
14
+ [2.00 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/2.0bpw)
15
+ [2.25 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/2.25bpw)
16
+ [2.50 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/2.5bpw)
17
+ [3.00 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/3.0bpw)
18
+ [3.50 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/3.5bpw)
19
+ [4.00 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/4.0bpw)
20
+ [5.00 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/5.0bpw)
21
+ [6.00 bits per weight](https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/tree/6.0bpw)
22
+
23
+ # SVG Catbench
24
+
25
+ <table>
26
+ <tr>
27
+ <td align="center">
28
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/2.0bpw.svg">
29
+ <img src="2.0bpw.svg" alt="2.00 bpw" width="160">
30
+ </a>
31
+ <div>2.00 bpw</div>
32
+ </td>
33
+ <td align="center">
34
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/2.25bpw.svg">
35
+ <img src="2.25bpw.svg" alt="2.25 bpw" width="160">
36
+ </a>
37
+ <div>2.25 bpw</div>
38
+ </td>
39
+ <td align="center">
40
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/2.5bpw.svg">
41
+ <img src="2.5bpw.svg" alt="2.5 bpw" width="160">
42
+ </a>
43
+ <div>2.5 bpw</div>
44
+ </td>
45
+ <td align="center">
46
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/3.0bpw.svg">
47
+ <img src="3.0bpw.svg" alt="3.00 bpw" width="160">
48
+ </a>
49
+ <div>3.00 bpw</div>
50
+ </td>
51
+ </tr>
52
+ <tr>
53
+ <td align="center">
54
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/3.5bpw.svg">
55
+ <img src="3.5bpw.svg" alt="3.50 bpw" width="160">
56
+ </a>
57
+ <div>3.50 bpw</div>
58
+ </td>
59
+ <td align="center">
60
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/4.0bpw.svg">
61
+ <img src="4.0bpw.svg" alt="4.00 bpw" width="160">
62
+ </a>
63
+ <div>4.00 bpw</div>
64
+ </td>
65
+ <td align="center">
66
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/5.0bpw.svg">
67
+ <img src="5.0bpw.svg" alt="5.00 bpw" width="160">
68
+ </a>
69
+ <div>5.00 bpw</div>
70
+ </td>
71
+ <td align="center">
72
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-32B-Instruct-exl3/blob/main/6.0bpw.svg">
73
+ <img src="6.0bpw.svg" alt="6.00 bpw" width="160">
74
+ </a>
75
+ <div>6.00 bpw</div>
76
+ </td>
77
+ </tr>
78
+ <tr>
79
+ <td align="center">
80
+ <a href="https://huggingface.co/turboderp/Qwen3-VL-8B-Instruct-exl3/blob/main/api.svg">
81
+ <img src="api.svg" alt="API" width="160">
82
+ </a>
83
+ <div>API</div>
84
+ </td>
85
+ </tr>
86
+ </table>