Hugging Face's logo

second-state
/
Llama-3-8B-Japanese-Instruct-GGUF

Text Generation
GGUF
English
Japanese
conversational
Model card Files Community
apepkuss79 commited on
Commit
51dd8d6
•
1 Parent(s): 9a5ca84

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +88 -1
README.md CHANGED
@@ -1,3 +1,90 @@
1
  ---
2
- license: apache-2.0
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: other
3
+ license_name: llama3
4
+ base_model: haqishen/Llama-3-8B-Japanese-Instruct
5
+ inference: false
6
+ model_creator: haqishen
7
+ model_type: llama
8
+ pipeline_tag: text-generation
9
+ quantized_by: Second State Inc.
10
+ language:
11
+ - en
12
+ - ja
13
  ---
14
+
15
+ <!-- header start -->
16
+ <!-- 200823 -->
17
+ <div style="width: auto; margin-left: auto; margin-right: auto">
18
+ <img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
19
+ </div>
20
+ <hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
21
+ <!-- header end -->
22
+
23
+ # Llama3-8B-Japanese-Chat-GGUF
24
+
25
+ ## Original Model
26
+
27
+ [haqishen/Llama-3-8B-Japanese-Instruct](https://huggingface.co/haqishen/Llama-3-8B-Japanese-Instruct)
28
+
29
+ ## Run with LlamaEdge
30
+
31
+ - LlamaEdge version: [v0.10.1](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.10.1) and above
32
+
33
+ - Prompt template
34
+
35
+ - Prompt type: `llama-3-chat`
36
+
37
+ - Prompt string
38
+
39
+ ```text
40
+ <|begin_of_text|><|start_header_id|>system<|end_header_id|>
41
+
42
+ {{ system_prompt }}<|eot_id|><|start_header_id|>user<|end_header_id|>
43
+
44
+ {{ user_message_1 }}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
45
+
46
+ {{ model_answer_1 }}<|eot_id|><|start_header_id|>user<|end_header_id|>
47
+
48
+ {{ user_message_2 }}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
49
+ ```
50
+
51
+ - Context size: `4096`
52
+
53
+ - Run as LlamaEdge service
54
+
55
+ ```bash
56
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:Llama-3-8B-Japanese-Instruct-Q5_K_M.gguf \
57
+ llama-api-server.wasm \
58
+ --prompt-template llama-3-chat \
59
+ --ctx-size 4096 \
60
+ --model-name Llama-3-8B-Japanese-Instruct \
61
+ ```
62
+
63
+ - Run as LlamaEdge command app
64
+
65
+ ```bash
66
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:Llama-3-8B-Japanese-Instruct-Q5_K_M.gguf \
67
+ llama-chat.wasm \
68
+ --prompt-template Llama-3-8B-Japanese-Instruct \
69
+ --ctx-size 4096 \
70
+ ```
71
+ <!--
72
+ ## Quantized GGUF Models
73
+
74
+ | Name | Quant method | Bits | Size | Use case |
75
+ | ---- | ---- | ---- | ---- | ----- |
76
+ | [Llama3-8B-Japanese-Chat-Q2_K.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q2_K.gguf) | Q2_K | 2 | 3.18 GB| smallest, significant quality loss - not recommended for most purposes |
77
+ | [Llama3-8B-Japanese-Chat-Q3_K_L.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q3_K_L.gguf) | Q3_K_L | 3 | 4.32 GB| small, substantial quality loss |
78
+ | [Llama3-8B-Japanese-Chat-Q3_K_M.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q3_K_M.gguf) | Q3_K_M | 3 | 4.02 GB| very small, high quality loss |
79
+ | [Llama3-8B-Japanese-Chat-Q3_K_S.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q3_K_S.gguf) | Q3_K_S | 3 | 3.66 GB| very small, high quality loss |
80
+ | [Llama3-8B-Japanese-Chat-Q4_0.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q4_0.gguf) | Q4_0 | 4 | 4.66 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
81
+ | [Llama3-8B-Japanese-Chat-Q4_K_M.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q4_K_M.gguf) | Q4_K_M | 4 | 4.92 GB| medium, balanced quality - recommended |
82
+ | [Llama3-8B-Japanese-Chat-Q4_K_S.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q4_K_S.gguf) | Q4_K_S | 4 | 4.69 GB| small, greater quality loss |
83
+ | [Llama3-8B-Japanese-Chat-Q5_0.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q5_0.gguf) | Q5_0 | 5 | 5.6 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
84
+ | [Llama3-8B-Japanese-Chat-Q5_K_M.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q5_K_M.gguf) | Q5_K_M | 5 | 5.73 GB| large, very low quality loss - recommended |
85
+ | [Llama3-8B-Japanese-Chat-Q5_K_S.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q5_K_S.gguf) | Q5_K_S | 5 | 5.6 GB| large, low quality loss - recommended |
86
+ | [Llama3-8B-Japanese-Chat-Q6_K.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q6_K.gguf) | Q6_K | 6 | 6.6 GB| very large, extremely low quality loss |
87
+ | [Llama3-8B-Japanese-Chat-Q8_0.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-Q8_0.gguf) | Q8_0 | 8 | 8.54 GB| very large, extremely low quality loss - not recommended |
88
+ | [Llama3-8B-Japanese-Chat-f16.gguf](https://huggingface.co/second-state/Llama3-8B-Japanese-Chat-GGUF/blob/main/Llama3-8B-Japanese-Chat-f16.gguf) | f16 | 16 | 16.1 GB| | -->
89
+
90
+ *Quantized with llama.cpp b2824.*