Hugging Face's logo

second-state
/
Phi-4-reasoning-GGUF

Text Generation
Transformers
GGUF
English
phi3
conversational
Model card Files Community
apepkuss79 commited on
Commit
2568ae4
·
verified ·
1 Parent(s): 2d2cef9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +87 -87
README.md CHANGED
@@ -1,88 +1,88 @@
1
- ---
2
- license: mit
3
- license_link: https://huggingface.co/microsoft/Phi-4-reasoning/resolve/main/LICENSE
4
- language:
5
- - en
6
- base_model:
7
- - microsoft/phi-4
8
- pipeline_tag: text-generation
9
- library_name: transformers
10
- model_creator: Microsoft
11
- model_name: Phi-4-reasoning
12
- quantized_by: Second State Inc.
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
- # Phi-4-reasoning-GGUF
24
-
25
- ## Original Model
26
-
27
- [microsoft/Phi-4-reasoning](https://huggingface.co/microsoft/Phi-4-reasoning)
28
-
29
- ## Run with LlamaEdge
30
-
31
- - LlamaEdge version: coming soon
32
-
33
- <!-- - LlamaEdge version: [v0.16.0](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.16.0) and above -->
34
-
35
- - Prompt template
36
-
37
- - Prompt type: `Phi-4-chat`
38
-
39
- - Prompt string
40
-
41
- ```text
42
- <|im_start|>system<|im_sep|>
43
- {system_message}<|im_end|>
44
- <|im_start|>user<|im_sep|>
45
- {user_message}<|im_end|>
46
- <|im_start|>assistant<|im_sep|>
47
- ```
48
-
49
- - Context size: `32000`
50
-
51
- - Run as LlamaEdge service
52
-
53
- ```bash
54
- wasmedge --dir .:. --nn-preload default:GGML:AUTO:Phi-4-reasoning-Q5_K_M.gguf \
55
- llama-api-server.wasm \
56
- --prompt-template Phi-4-chat \
57
- --ctx-size 32000 \
58
- --model-name phi-4-reasoning
59
- ```
60
-
61
- - Run as LlamaEdge command app
62
-
63
- ```bash
64
- wasmedge --dir .:. --nn-preload default:GGML:AUTO:Phi-4-reasoning-Q5_K_M.gguf \
65
- llama-chat.wasm \
66
- --prompt-template Phi-4-chat \
67
- --ctx-size 32000
68
- ```
69
-
70
- ## Quantized GGUF Models
71
-
72
- | Name | Quant method | Bits | Size | Use case |
73
- | ---- | ---- | ---- | ---- | ----- |
74
- | [Phi-4-reasoning-Q2_K.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q2_K.gguf) | Q2_K | 2 | 5.55 GB| smallest, significant quality loss - not recommended for most purposes |
75
- | [Phi-4-reasoning-Q3_K_L.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q3_K_L.gguf) | Q3_K_L | 3 | 7.93 GB| small, substantial quality loss |
76
- | [Phi-4-reasoning-Q3_K_M.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q3_K_M.gguf) | Q3_K_M | 3 | 7.36 GB| very small, high quality loss |
77
- | [Phi-4-reasoning-Q3_K_S.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q3_K_S.gguf) | Q3_K_S | 3 | 6.50 GB| very small, high quality loss |
78
- | [Phi-4-reasoning-Q4_0.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q4_0.gguf) | Q4_0 | 4 | 8.38 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
79
- | [Phi-4-reasoning-Q4_K_M.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q4_K_M.gguf) | Q4_K_M | 4 | 9.05 GB| medium, balanced quality - recommended |
80
- | [Phi-4-reasoning-Q4_K_S.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q4_K_S.gguf) | Q4_K_S | 4 | 8.44 GB| small, greater quality loss |
81
- | [Phi-4-reasoning-Q5_0.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q5_0.gguf) | Q5_0 | 5 | 10.2 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
82
- | [Phi-4-reasoning-Q5_K_M.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q5_K_M.gguf) | Q5_K_M | 5 | 10.6 GB| large, very low quality loss - recommended |
83
- | [Phi-4-reasoning-Q5_K_S.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q5_K_S.gguf) | Q5_K_S | 5 | 10.2 GB| large, low quality loss - recommended |
84
- | [Phi-4-reasoning-Q6_K.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q6_K.gguf) | Q6_K | 6 | 12.0 GB| very large, extremely low quality loss |
85
- | [Phi-4-reasoning-Q8_0.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q8_0.gguf) | Q8_0 | 8 | 15.6 GB| very large, extremely low quality loss - not recommended |
86
- | [Phi-4-reasoning-f16.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-f16.gguf) | f16 | 16 | 29.3 GB| |
87
-
88
  *Quantized with llama.cpp b5201.*
 
1
+ ---
2
+ license: mit
3
+ license_link: https://huggingface.co/microsoft/Phi-4-reasoning/resolve/main/LICENSE
4
+ language:
5
+ - en
6
+ base_model:
7
+ - microsoft/phi-4
8
+ pipeline_tag: text-generation
9
+ library_name: transformers
10
+ model_creator: Microsoft
11
+ model_name: Phi-4-reasoning
12
+ quantized_by: Second State Inc.
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
+ # Phi-4-reasoning-GGUF
24
+
25
+ ## Original Model
26
+
27
+ [microsoft/Phi-4-reasoning](https://huggingface.co/microsoft/Phi-4-reasoning)
28
+
29
+ ## Run with LlamaEdge
30
+
31
+ - LlamaEdge version: coming soon
32
+
33
+ <!-- - LlamaEdge version: [v0.16.0](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.16.0) and above -->
34
+
35
+ - Prompt template
36
+
37
+ - Prompt type: `phi-4-chat`
38
+
39
+ - Prompt string
40
+
41
+ ```text
42
+ <|im_start|>system<|im_sep|>
43
+ {system_message}<|im_end|>
44
+ <|im_start|>user<|im_sep|>
45
+ {user_message}<|im_end|>
46
+ <|im_start|>assistant<|im_sep|>
47
+ ```
48
+
49
+ - Context size: `32000`
50
+
51
+ - Run as LlamaEdge service
52
+
53
+ ```bash
54
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:Phi-4-reasoning-Q5_K_M.gguf \
55
+ llama-api-server.wasm \
56
+ --prompt-template phi-4-chat \
57
+ --ctx-size 32000 \
58
+ --model-name phi-4-reasoning
59
+ ```
60
+
61
+ - Run as LlamaEdge command app
62
+
63
+ ```bash
64
+ wasmedge --dir .:. --nn-preload default:GGML:AUTO:Phi-4-reasoning-Q5_K_M.gguf \
65
+ llama-chat.wasm \
66
+ --prompt-template phi-4-chat \
67
+ --ctx-size 32000
68
+ ```
69
+
70
+ ## Quantized GGUF Models
71
+
72
+ | Name | Quant method | Bits | Size | Use case |
73
+ | ---- | ---- | ---- | ---- | ----- |
74
+ | [Phi-4-reasoning-Q2_K.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q2_K.gguf) | Q2_K | 2 | 5.55 GB| smallest, significant quality loss - not recommended for most purposes |
75
+ | [Phi-4-reasoning-Q3_K_L.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q3_K_L.gguf) | Q3_K_L | 3 | 7.93 GB| small, substantial quality loss |
76
+ | [Phi-4-reasoning-Q3_K_M.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q3_K_M.gguf) | Q3_K_M | 3 | 7.36 GB| very small, high quality loss |
77
+ | [Phi-4-reasoning-Q3_K_S.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q3_K_S.gguf) | Q3_K_S | 3 | 6.50 GB| very small, high quality loss |
78
+ | [Phi-4-reasoning-Q4_0.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q4_0.gguf) | Q4_0 | 4 | 8.38 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
79
+ | [Phi-4-reasoning-Q4_K_M.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q4_K_M.gguf) | Q4_K_M | 4 | 9.05 GB| medium, balanced quality - recommended |
80
+ | [Phi-4-reasoning-Q4_K_S.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q4_K_S.gguf) | Q4_K_S | 4 | 8.44 GB| small, greater quality loss |
81
+ | [Phi-4-reasoning-Q5_0.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q5_0.gguf) | Q5_0 | 5 | 10.2 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
82
+ | [Phi-4-reasoning-Q5_K_M.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q5_K_M.gguf) | Q5_K_M | 5 | 10.6 GB| large, very low quality loss - recommended |
83
+ | [Phi-4-reasoning-Q5_K_S.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q5_K_S.gguf) | Q5_K_S | 5 | 10.2 GB| large, low quality loss - recommended |
84
+ | [Phi-4-reasoning-Q6_K.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q6_K.gguf) | Q6_K | 6 | 12.0 GB| very large, extremely low quality loss |
85
+ | [Phi-4-reasoning-Q8_0.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-Q8_0.gguf) | Q8_0 | 8 | 15.6 GB| very large, extremely low quality loss - not recommended |
86
+ | [Phi-4-reasoning-f16.gguf](https://huggingface.co/second-state/Phi-4-reasoning-GGUF/blob/main/Phi-4-reasoning-f16.gguf) | f16 | 16 | 29.3 GB| |
87
+
88
  *Quantized with llama.cpp b5201.*