karimouda commited on
Commit
9596bb1
·
verified ·
1 Parent(s): 0b470e4

Add results file for phi-4

Browse files
results/microsoft/phi-4_results_2025-05-10_16-20-07.json ADDED
@@ -0,0 +1,153 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "results": {
3
+ "average_score": 7.304255319148936,
4
+ "speed": 4.860072067723493,
5
+ "contamination_score": 0,
6
+ "execution_time": 4194.382247,
7
+ "errors": [],
8
+ "scores_by_category": [
9
+ {
10
+ "category": "Paraphrasing",
11
+ "average_score": 10.0,
12
+ "count": 6
13
+ },
14
+ {
15
+ "category": "Reading Comprehension",
16
+ "average_score": 10.0,
17
+ "count": 17
18
+ },
19
+ {
20
+ "category": "Coding",
21
+ "average_score": 10.0,
22
+ "count": 3
23
+ },
24
+ {
25
+ "category": "Entity Extraction",
26
+ "average_score": 9.0,
27
+ "count": 5
28
+ },
29
+ {
30
+ "category": "Long Context",
31
+ "average_score": 9.0,
32
+ "count": 4
33
+ },
34
+ {
35
+ "category": "Sentiment Analysis",
36
+ "average_score": 8.88888888888889,
37
+ "count": 9
38
+ },
39
+ {
40
+ "category": "General Knowledge",
41
+ "average_score": 8.095238095238095,
42
+ "count": 63
43
+ },
44
+ {
45
+ "category": "MMLU",
46
+ "average_score": 8.016528925619834,
47
+ "count": 121
48
+ },
49
+ {
50
+ "category": "Trust & Safety",
51
+ "average_score": 8.0,
52
+ "count": 30
53
+ },
54
+ {
55
+ "category": "RAG QA",
56
+ "average_score": 7.658536585365853,
57
+ "count": 41
58
+ },
59
+ {
60
+ "category": "Reasoning & Math",
61
+ "average_score": 7.046511627906977,
62
+ "count": 43
63
+ },
64
+ {
65
+ "category": "Transliteration",
66
+ "average_score": 6.833333333333333,
67
+ "count": 6
68
+ },
69
+ {
70
+ "category": "Summarization",
71
+ "average_score": 6.75,
72
+ "count": 8
73
+ },
74
+ {
75
+ "category": "Writing (incl Dialects)",
76
+ "average_score": 6.545454545454546,
77
+ "count": 22
78
+ },
79
+ {
80
+ "category": "Diacritization",
81
+ "average_score": 6.083333333333333,
82
+ "count": 12
83
+ },
84
+ {
85
+ "category": "Structuring",
86
+ "average_score": 6.0,
87
+ "count": 3
88
+ },
89
+ {
90
+ "category": "Function Calling",
91
+ "average_score": 6.0,
92
+ "count": 3
93
+ },
94
+ {
95
+ "category": "Translation (incl Dialects)",
96
+ "average_score": 5.666666666666667,
97
+ "count": 36
98
+ },
99
+ {
100
+ "category": "Arabic Language & Grammar",
101
+ "average_score": 4.0,
102
+ "count": 17
103
+ },
104
+ {
105
+ "category": "Instruction Following",
106
+ "average_score": 3.7142857142857144,
107
+ "count": 7
108
+ },
109
+ {
110
+ "category": "Dialect Detection",
111
+ "average_score": 2.6363636363636362,
112
+ "count": 11
113
+ },
114
+ {
115
+ "category": "Hallucination",
116
+ "average_score": 0.0,
117
+ "count": 3
118
+ }
119
+ ],
120
+ "scores_by_format": [
121
+ {
122
+ "format": "MCQ",
123
+ "average_score": 8.25764192139738,
124
+ "count": 229
125
+ },
126
+ {
127
+ "format": "Short Answer",
128
+ "average_score": 8.0,
129
+ "count": 5
130
+ },
131
+ {
132
+ "format": "Generation",
133
+ "average_score": 6.456140350877193,
134
+ "count": 228
135
+ },
136
+ {
137
+ "format": "Fill-in-the-blank",
138
+ "average_score": 3.75,
139
+ "count": 8
140
+ }
141
+ ]
142
+ },
143
+ "config": {
144
+ "model": "microsoft/phi-4",
145
+ "model_sha": "187ef0342fff0eb3333be9f00389385e95ef0b61",
146
+ "submitted_time": "2025-05-10T14:54:51Z",
147
+ "likes": 2029,
148
+ "params": 14.66,
149
+ "license": "mit",
150
+ "model_source": "Hugging Face",
151
+ "model_category": "Medium"
152
+ }
153
+ }