3morixd commited on
Commit
56a46a3
·
verified ·
1 Parent(s): 609be0b

Upload ranking.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. ranking.json +250 -0
ranking.json ADDED
@@ -0,0 +1,250 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "rank": 1,
4
+ "model": "SmolLM2-135M-Instruct-mobile",
5
+ "cpu_tps": 59.7,
6
+ "size_mb": 1,
7
+ "efficiency_score": 59.7,
8
+ "verdict": "BEST"
9
+ },
10
+ {
11
+ "rank": 2,
12
+ "model": "Qwen2.5-0.5B-Instruct-mobile-int4",
13
+ "cpu_tps": 12.5,
14
+ "size_mb": 1,
15
+ "efficiency_score": 12.5,
16
+ "verdict": "BEST"
17
+ },
18
+ {
19
+ "rank": 3,
20
+ "model": "Llama-3.2-1B-Instruct-Q4-mobile",
21
+ "cpu_tps": 11.3,
22
+ "size_mb": 1,
23
+ "efficiency_score": 11.3,
24
+ "verdict": "BEST"
25
+ },
26
+ {
27
+ "rank": 4,
28
+ "model": "SmolLM2-360M-Instruct-mobile",
29
+ "cpu_tps": 29.1,
30
+ "size_mb": 258.1,
31
+ "efficiency_score": 0.113,
32
+ "verdict": "GOOD"
33
+ },
34
+ {
35
+ "rank": 5,
36
+ "model": "MiniCPM-V-4.6-mobile",
37
+ "cpu_tps": 23.4,
38
+ "size_mb": 504.6,
39
+ "efficiency_score": 0.046,
40
+ "verdict": "GOOD"
41
+ },
42
+ {
43
+ "rank": 6,
44
+ "model": "Qwen2.5-0.5B-Chinese-mobile",
45
+ "cpu_tps": 16.4,
46
+ "size_mb": 400.6,
47
+ "efficiency_score": 0.041,
48
+ "verdict": "GOOD"
49
+ },
50
+ {
51
+ "rank": 7,
52
+ "model": "Qwen2.5-0.5B-Instruct-Server-mobile",
53
+ "cpu_tps": 16.7,
54
+ "size_mb": 482.3,
55
+ "efficiency_score": 0.035,
56
+ "verdict": "GOOD"
57
+ },
58
+ {
59
+ "rank": 8,
60
+ "model": "MiniCPM5-1B-multilingual-mobile",
61
+ "cpu_tps": 18.4,
62
+ "size_mb": 656.2,
63
+ "efficiency_score": 0.028,
64
+ "verdict": "GOOD"
65
+ },
66
+ {
67
+ "rank": 9,
68
+ "model": "TinyLlama-1.1B-Chat-mobile-int4",
69
+ "cpu_tps": 17.6,
70
+ "size_mb": 637.8,
71
+ "efficiency_score": 0.028,
72
+ "verdict": "GOOD"
73
+ },
74
+ {
75
+ "rank": 10,
76
+ "model": "MiniCPM5-1B-mobile",
77
+ "cpu_tps": 18.1,
78
+ "size_mb": 656.2,
79
+ "efficiency_score": 0.028,
80
+ "verdict": "GOOD"
81
+ },
82
+ {
83
+ "rank": 11,
84
+ "model": "Qwen2.5-0.5B-Coder-mobile",
85
+ "cpu_tps": 8.0,
86
+ "size_mb": 379.4,
87
+ "efficiency_score": 0.021,
88
+ "verdict": "OK"
89
+ },
90
+ {
91
+ "rank": 12,
92
+ "model": "SmolLM2-1.7B-Instruct-mobile",
93
+ "cpu_tps": 18.7,
94
+ "size_mb": 1006.7,
95
+ "efficiency_score": 0.019,
96
+ "verdict": "OK"
97
+ },
98
+ {
99
+ "rank": 13,
100
+ "model": "Qwen2.5-Math-1.5B-mobile",
101
+ "cpu_tps": 15.7,
102
+ "size_mb": 940.4,
103
+ "efficiency_score": 0.017,
104
+ "verdict": "OK"
105
+ },
106
+ {
107
+ "rank": 14,
108
+ "model": "TinyLlama-1.1B-Chat-Q5-mobile",
109
+ "cpu_tps": 10.9,
110
+ "size_mb": 746.7,
111
+ "efficiency_score": 0.015,
112
+ "verdict": "OK"
113
+ },
114
+ {
115
+ "rank": 15,
116
+ "model": "Gemma-2B-Arabic-mobile",
117
+ "cpu_tps": 14.8,
118
+ "size_mb": 1554.7,
119
+ "efficiency_score": 0.01,
120
+ "verdict": "OK"
121
+ },
122
+ {
123
+ "rank": 16,
124
+ "model": "SmolLM2-1.7B-Instruct-Q5-mobile",
125
+ "cpu_tps": 10.8,
126
+ "size_mb": 1168.7,
127
+ "efficiency_score": 0.009,
128
+ "verdict": "OK"
129
+ },
130
+ {
131
+ "rank": 17,
132
+ "model": "Qwen2.5-Coder-1.5B-mobile",
133
+ "cpu_tps": 7.8,
134
+ "size_mb": 940.4,
135
+ "efficiency_score": 0.008,
136
+ "verdict": "OK"
137
+ },
138
+ {
139
+ "rank": 18,
140
+ "model": "Qwen2.5-1.5B-Instruct-mobile-int4",
141
+ "cpu_tps": 7.7,
142
+ "size_mb": 1065.6,
143
+ "efficiency_score": 0.007,
144
+ "verdict": "OK"
145
+ },
146
+ {
147
+ "rank": 19,
148
+ "model": "Qwen2.5-1.5B-Instruct-Q8-mobile",
149
+ "cpu_tps": 8.7,
150
+ "size_mb": 1570.3,
151
+ "efficiency_score": 0.006,
152
+ "verdict": "OK"
153
+ },
154
+ {
155
+ "rank": 20,
156
+ "model": "Llama-3.2-1B-Instruct-Q6-mobile",
157
+ "cpu_tps": 5.0,
158
+ "size_mb": 974.5,
159
+ "efficiency_score": 0.005,
160
+ "verdict": "OK"
161
+ },
162
+ {
163
+ "rank": 21,
164
+ "model": "Gemma-2-2B-IT-mobile",
165
+ "cpu_tps": 8.2,
166
+ "size_mb": 1629.4,
167
+ "efficiency_score": 0.005,
168
+ "verdict": "OK"
169
+ },
170
+ {
171
+ "rank": 22,
172
+ "model": "Llama-3.2-3B-FunctionCall-mobile",
173
+ "cpu_tps": 9.1,
174
+ "size_mb": 1925.8,
175
+ "efficiency_score": 0.005,
176
+ "verdict": "OK"
177
+ },
178
+ {
179
+ "rank": 23,
180
+ "model": "Llama-3.2-1B-FunctionCall-mobile",
181
+ "cpu_tps": 8.9,
182
+ "size_mb": 1925.8,
183
+ "efficiency_score": 0.005,
184
+ "verdict": "OK"
185
+ },
186
+ {
187
+ "rank": 24,
188
+ "model": "Qwen2.5-1.5B-Instruct-Q5-mobile",
189
+ "cpu_tps": 4.4,
190
+ "size_mb": 1072.9,
191
+ "efficiency_score": 0.004,
192
+ "verdict": "OK"
193
+ },
194
+ {
195
+ "rank": 25,
196
+ "model": "Llama-3.2-1B-Instruct-mobile",
197
+ "cpu_tps": 4.9,
198
+ "size_mb": 1259.9,
199
+ "efficiency_score": 0.004,
200
+ "verdict": "OK"
201
+ },
202
+ {
203
+ "rank": 26,
204
+ "model": "Phi-3.5-mini-Instruct-mobile",
205
+ "cpu_tps": 8.6,
206
+ "size_mb": 2282.4,
207
+ "efficiency_score": 0.004,
208
+ "verdict": "OK"
209
+ },
210
+ {
211
+ "rank": 27,
212
+ "model": "Gemma-2-2B-IT-Q5-mobile",
213
+ "cpu_tps": 5.7,
214
+ "size_mb": 1834.2,
215
+ "efficiency_score": 0.003,
216
+ "verdict": "OK"
217
+ },
218
+ {
219
+ "rank": 28,
220
+ "model": "Phi-3.5-mini-instruct-Q5-mobile",
221
+ "cpu_tps": 7.0,
222
+ "size_mb": 2684.9,
223
+ "efficiency_score": 0.003,
224
+ "verdict": "OK"
225
+ },
226
+ {
227
+ "rank": 29,
228
+ "model": "Llama-3.2-3B-Instruct-mobile",
229
+ "cpu_tps": 4.8,
230
+ "size_mb": 1925.8,
231
+ "efficiency_score": 0.002,
232
+ "verdict": "OK"
233
+ },
234
+ {
235
+ "rank": 30,
236
+ "model": "Llama-3.2-3B-Instruct-Q5-mobile",
237
+ "cpu_tps": 2.6,
238
+ "size_mb": 2214.6,
239
+ "efficiency_score": 0.001,
240
+ "verdict": "OK"
241
+ },
242
+ {
243
+ "rank": 31,
244
+ "model": "Qwen2.5-Coder-7B-mobile",
245
+ "cpu_tps": 3.0,
246
+ "size_mb": 4466.1,
247
+ "efficiency_score": 0.001,
248
+ "verdict": "OK"
249
+ }
250
+ ]