Hugging Face's logo

second-state
/
Qwen2.5-3B-Instruct-GGUF

Text Generation
GGUF
English
qwen2
chat
conversational
Model card Files
xet
 Community
1

Improve language tag

#1
by lbourdois - opened
base: refs/heads/main
←
from: refs/pr/1
Discussion Files changed
+99
-87
Files changed (1) hide show
  1. README.md +99 -87
README.md CHANGED
@@ -1,87 +1,99 @@
1
- ---
2
- base_model: Qwen/Qwen2.5-3B-Instruct
3
- license: other
4
- license_name: qwen-research
5
- license_link: https://huggingface.co/Qwen/Qwen2.5-3B-Instruct/blob/main/LICENSEE
6
- model_creator: Qwen
7
- model_name: Qwen2.5-3B-Instruct
8
- quantized_by: Second State Inc.
9
- language:
10
- - en
11
- pipeline_tag: text-generation
12
- tags:
13
- - chat
14
- ---
15
-
16
- <!-- header start -->
17
- <!-- 200823 -->
18
- <div style="width: auto; margin-left: auto; margin-right: auto">
19
- <img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
20
- </div>
21
- <hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
22
- <!-- header end -->
23
-
24
- # Qwen2.5-3B-Instruct-GGUF
25
-
26
- ## Original Model
27
-
28
- [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct)
29
-
30
- ## Run with LlamaEdge
31
-
32
- - LlamaEdge version: [v0.14.3](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.14.3) and above
33
-
34
- - Prompt template
35
-
36
- - Prompt type: `chatml`
37
-
38
- - Prompt string
39
-
40
- ```text
41
- <|im_start|>system
42
- {system_message}<|im_end|>
43
- <|im_start|>user
44
- {prompt}<|im_end|>
45
- <|im_start|>assistant
46
- ```
47
-
48
- - Context size: `32000`
49
-
50
- - Run as LlamaEdge service
51
-
52
- ```bash
53
- wasmedge --dir .:. --nn-preload default:GGML:AUTO:Qwen2.5-3B-Instruct-Q5_K_M.gguf \
54
- llama-api-server.wasm \
55
- --model-name Qwen2.5-3B-Instruct \
56
- --prompt-template chatml \
57
- --ctx-size 32000
58
- ```
59
-
60
- - Run as LlamaEdge command app
61
-
62
- ```bash
63
- wasmedge --dir .:. --nn-preload default:GGML:AUTO:Qwen2.5-3B-Instruct-Q5_K_M.gguf \
64
- llama-chat.wasm \
65
- --prompt-template chatml \
66
- --ctx-size 32000
67
- ```
68
-
69
- ## Quantized GGUF Models
70
-
71
- | Name | Quant method | Bits | Size | Use case |
72
- | ---- | ---- | ---- | ---- | ----- |
73
- | [Qwen2.5-3B-Instruct-Q2_K.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q2_K.gguf) | Q2_K | 2 | 676 MB| smallest, significant quality loss - not recommended for most purposes |
74
- | [Qwen2.5-3B-Instruct-Q3_K_L.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q3_K_L.gguf) | Q3_K_L | 3 | 880 MB| small, substantial quality loss |
75
- | [Qwen2.5-3B-Instruct-Q3_K_M.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q3_K_M.gguf) | Q3_K_M | 3 | 824 MB| very small, high quality loss |
76
- | [Qwen2.5-3B-Instruct-Q3_K_S.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q3_K_S.gguf) | Q3_K_S | 3 | 761 MB| very small, high quality loss |
77
- | [Qwen2.5-3B-Instruct-Q4_0.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q4_0.gguf) | Q4_0 | 4 | 935 MB| legacy; small, very high quality loss - prefer using Q3_K_M |
78
- | [Qwen2.5-3B-Instruct-Q4_K_M.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q4_K_M.gguf) | Q4_K_M | 4 | 986 MB| medium, balanced quality - recommended |
79
- | [Qwen2.5-3B-Instruct-Q4_K_S.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q4_K_S.gguf) | Q4_K_S | 4 | 940 MB| small, greater quality loss |
80
- | [Qwen2.5-3B-Instruct-Q5_0.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q5_0.gguf) | Q5_0 | 5 | 1.1 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
81
- | [Qwen2.5-3B-Instruct-Q5_K_M.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q5_K_M.gguf) | Q5_K_M | 5 | 1.13 GB| large, very low quality loss - recommended |
82
- | [Qwen2.5-3B-Instruct-Q5_K_S.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q5_K_S.gguf) | Q5_K_S | 5 | 1.1 GB| large, low quality loss - recommended |
83
- | [Qwen2.5-3B-Instruct-Q6_K.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q6_K.gguf) | Q6_K | 6 | 1.27 GB| very large, extremely low quality loss |
84
- | [Qwen2.5-3B-Instruct-Q8_0.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q8_0.gguf) | Q8_0 | 8 | 1.65 GB| very large, extremely low quality loss - not recommended |
85
- | [Qwen2.5-3B-Instruct-f16.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-f16.gguf) | f16 | 16 | 3.09 GB| |
86
-
87
- *Quantized with llama.cpp b4466*
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Qwen/Qwen2.5-3B-Instruct
3
+ license: other
4
+ license_name: qwen-research
5
+ license_link: https://huggingface.co/Qwen/Qwen2.5-3B-Instruct/blob/main/LICENSEE
6
+ model_creator: Qwen
7
+ model_name: Qwen2.5-3B-Instruct
8
+ quantized_by: Second State Inc.
9
+ language:
10
+ - zho
11
+ - eng
12
+ - fra
13
+ - spa
14
+ - por
15
+ - deu
16
+ - ita
17
+ - rus
18
+ - jpn
19
+ - kor
20
+ - vie
21
+ - tha
22
+ - ara
23
+ pipeline_tag: text-generation
24
+ tags:
25
+ - chat
26
+ ---
27
+
28
+ <!-- header start -->
29
+ <!-- 200823 -->
30
+ <div style="width: auto; margin-left: auto; margin-right: auto">
31
+ <img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
32
+ </div>
33
+ <hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
34
+ <!-- header end -->
35
+
36
+ # Qwen2.5-3B-Instruct-GGUF
37
+
38
+ ## Original Model
39
+
40
+ [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct)
41
+
42
+ ## Run with LlamaEdge
43
+
44
+ - LlamaEdge version: [v0.14.3](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.14.3) and above
45
+
46
+ - Prompt template
47
+
48
+ - Prompt type: `chatml`
49
+
50
+ - Prompt string
51
+
52
+ ```text
53
+ <|im_start|>system
54
+ {system_message}<|im_end|>
55
+ <|im_start|>user
56
+ {prompt}<|im_end|>
57
+ <|im_start|>assistant
58
+ ```
59
+
60
+ - Context size: `32000`
61
+
62
+ - Run as LlamaEdge service
63
+
64
+ ```bash
65
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:Qwen2.5-3B-Instruct-Q5_K_M.gguf \
66
+ llama-api-server.wasm \
67
+ --model-name Qwen2.5-3B-Instruct \
68
+ --prompt-template chatml \
69
+ --ctx-size 32000
70
+ ```
71
+
72
+ - Run as LlamaEdge command app
73
+
74
+ ```bash
75
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:Qwen2.5-3B-Instruct-Q5_K_M.gguf \
76
+ llama-chat.wasm \
77
+ --prompt-template chatml \
78
+ --ctx-size 32000
79
+ ```
80
+
81
+ ## Quantized GGUF Models
82
+
83
+ | Name | Quant method | Bits | Size | Use case |
84
+ | ---- | ---- | ---- | ---- | ----- |
85
+ | [Qwen2.5-3B-Instruct-Q2_K.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q2_K.gguf) | Q2_K | 2 | 676 MB| smallest, significant quality loss - not recommended for most purposes |
86
+ | [Qwen2.5-3B-Instruct-Q3_K_L.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q3_K_L.gguf) | Q3_K_L | 3 | 880 MB| small, substantial quality loss |
87
+ | [Qwen2.5-3B-Instruct-Q3_K_M.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q3_K_M.gguf) | Q3_K_M | 3 | 824 MB| very small, high quality loss |
88
+ | [Qwen2.5-3B-Instruct-Q3_K_S.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q3_K_S.gguf) | Q3_K_S | 3 | 761 MB| very small, high quality loss |
89
+ | [Qwen2.5-3B-Instruct-Q4_0.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q4_0.gguf) | Q4_0 | 4 | 935 MB| legacy; small, very high quality loss - prefer using Q3_K_M |
90
+ | [Qwen2.5-3B-Instruct-Q4_K_M.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q4_K_M.gguf) | Q4_K_M | 4 | 986 MB| medium, balanced quality - recommended |
91
+ | [Qwen2.5-3B-Instruct-Q4_K_S.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q4_K_S.gguf) | Q4_K_S | 4 | 940 MB| small, greater quality loss |
92
+ | [Qwen2.5-3B-Instruct-Q5_0.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q5_0.gguf) | Q5_0 | 5 | 1.1 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
93
+ | [Qwen2.5-3B-Instruct-Q5_K_M.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q5_K_M.gguf) | Q5_K_M | 5 | 1.13 GB| large, very low quality loss - recommended |
94
+ | [Qwen2.5-3B-Instruct-Q5_K_S.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q5_K_S.gguf) | Q5_K_S | 5 | 1.1 GB| large, low quality loss - recommended |
95
+ | [Qwen2.5-3B-Instruct-Q6_K.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q6_K.gguf) | Q6_K | 6 | 1.27 GB| very large, extremely low quality loss |
96
+ | [Qwen2.5-3B-Instruct-Q8_0.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-Q8_0.gguf) | Q8_0 | 8 | 1.65 GB| very large, extremely low quality loss - not recommended |
97
+ | [Qwen2.5-3B-Instruct-f16.gguf](https://huggingface.co/second-state/Qwen2.5-3B-Instruct-GGUF/blob/main/Qwen2.5-3B-Instruct-f16.gguf) | f16 | 16 | 3.09 GB| |
98
+
99
+ *Quantized with llama.cpp b4466*