Hugging Face's logo

second-state
/
EXAONE-3.5-32B-Instruct-GGUF

Text Generation
Transformers
GGUF
English
Korean
exaone
lg-ai
exaone-3.5
custom_code
conversational
Model card Files Community
apepkuss79 commited on
Commit
b19260f
·
verified ·
1 Parent(s): 7eb7aea

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +92 -1
README.md CHANGED
@@ -1,3 +1,94 @@
1
  ---
2
- license: apache-2.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: other
3
+ license_name: exaone
4
+ license_link: LICENSE
5
+ model_name: EXAONE-3.5-32B-Instruct
6
+ base_model: LGAI-EXAONE/EXAONE-3.5-32B-Instruct
7
+ quantized_by: Second State Inc.
8
+ language:
9
+ - en
10
+ - ko
11
+ tags:
12
+ - lg-ai
13
+ - exaone
14
+ - exaone-3.5
15
+ pipeline_tag: text-generation
16
+ library_name: transformers
17
  ---
18
+
19
+ <!-- header start -->
20
+ <!-- 200823 -->
21
+ <div style="width: auto; margin-left: auto; margin-right: auto">
22
+ <img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
23
+ </div>
24
+ <hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
25
+ <!-- header end -->
26
+
27
+ # EXAONE-3.5-32B-Instruct-GGUF
28
+
29
+ ## Original Model
30
+
31
+ [LGAI-EXAONE/EXAONE-3.5-32B-Instruct](https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-32B-Instruct)
32
+
33
+ ## Run with LlamaEdge
34
+
35
+ - LlamaEdge version: coming soon
36
+
37
+ <!-- - LlamaEdge version: [v0.12.4](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.12.4) and above -->
38
+
39
+ - Prompt template
40
+
41
+ - Prompt type: `exaone-chat`
42
+
43
+ - Prompt string
44
+
45
+ ```text
46
+ [|system|]{system_message}[|endofturn|]
47
+ [|user|]{user_message_1}
48
+ [|assistant|]{assistant_message_1}[|endofturn|]
49
+ [|user|]{user_message_2}
50
+ [|assistant|]
51
+ ```
52
+
53
+ - Context size: `32000`
54
+
55
+ - Run as LlamaEdge service
56
+
57
+ ```bash
58
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:EXAONE-3.5-32B-Instruct-Q5_K_M.gguf \
59
+ llama-api-server.wasm \
60
+ --prompt-template exaone-chat \
61
+ --ctx-size 32000 \
62
+ --model-name EXAONE-3.5-32B-Instruct
63
+ ```
64
+
65
+ - Run as LlamaEdge command app
66
+
67
+ ```bash
68
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:EXAONE-3.5-32B-Instruct-Q5_K_M.gguf \
69
+ llama-chat.wasm \
70
+ --prompt-template exaone-chat \
71
+ --ctx-size 32000
72
+ ```
73
+
74
+ ## Quantized GGUF Models
75
+
76
+ | Name | Quant method | Bits | Size | Use case |
77
+ | ---- | ---- | ---- | ---- | ----- |
78
+ | [EXAONE-3.5-32B-Instruct-Q2_K.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q2_K.gguf) | Q2_K | 2 | 11.9 GB| smallest, significant quality loss - not recommended for most purposes |
79
+ | [EXAONE-3.5-32B-Instruct-Q3_K_L.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q3_K_L.gguf) | Q3_K_L | 3 | 16.8 GB| small, substantial quality loss |
80
+ | [EXAONE-3.5-32B-Instruct-Q3_K_M.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q3_K_M.gguf) | Q3_K_M | 3 | 15.5 GB| very small, high quality loss |
81
+ | [EXAONE-3.5-32B-Instruct-Q3_K_S.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q3_K_S.gguf) | Q3_K_S | 3 | 14.0 GB| very small, high quality loss |
82
+ | [EXAONE-3.5-32B-Instruct-Q4_0.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q4_0.gguf) | Q4_0 | 4 | 18.1 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
83
+ | [EXAONE-3.5-32B-Instruct-Q4_K_M.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q4_K_M.gguf) | Q4_K_M | 4 | 19.3 GB| medium, balanced quality - recommended |
84
+ | [EXAONE-3.5-32B-Instruct-Q4_K_S.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q4_K_S.gguf) | Q4_K_S | 4 | 18.3 GB| small, greater quality loss |
85
+ | [EXAONE-3.5-32B-Instruct-Q5_0.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q5_0.gguf) | Q5_0 | 5 | 22.1 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
86
+ | [EXAONE-3.5-32B-Instruct-Q5_K_M.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q5_K_M.gguf) | Q5_K_M | 5 | 22.7 GB| large, very low quality loss - recommended |
87
+ | [EXAONE-3.5-32B-Instruct-Q5_K_S.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q5_K_S.gguf) | Q5_K_S | 5 | 22.1 GB| large, low quality loss - recommended |
88
+ | [EXAONE-3.5-32B-Instruct-Q6_K.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q6_K.gguf) | Q6_K | 6 | 26.3 GB| very large, extremely low quality loss |
89
+ | [EXAONE-3.5-32B-Instruct-Q8_0.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-Q8_0.gguf) | Q8_0 | 8 | 34.0 GB| very large, extremely low quality loss - not recommended |
90
+ | [EXAONE-3.5-32B-Instruct-f16-00001-of-00003.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-f16-00001-of-00003.gguf) | f16 | 16 | 29.8 GB| |
91
+ | [EXAONE-3.5-32B-Instruct-f16-00002-of-00003.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-f16-00002-of-00003.gguf) | f16 | 16 | 30.0 GB| |
92
+ | [EXAONE-3.5-32B-Instruct-f16-00003-of-00003.gguf](https://huggingface.co/second-state/EXAONE-3.5-32B-Instruct-GGUF/blob/main/EXAONE-3.5-32B-Instruct-f16-00003-of-00003.gguf) | f16 | 16 | 4.23 GB| |
93
+
94
+ *Quantized with llama.cpp b4932.*