Spaces:
Running
Running
Update small.json
Browse files- small.json +93 -92
small.json
CHANGED
@@ -1,85 +1,94 @@
|
|
1 |
[
|
2 |
-
|
3 |
-
"model": "
|
4 |
-
"Average":
|
5 |
-
"MMLU":
|
6 |
-
"
|
7 |
-
"
|
8 |
-
"
|
9 |
-
"
|
10 |
-
"
|
11 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
},
|
|
|
13 |
{
|
14 |
-
"model": "SlimPajama",
|
15 |
-
"Average": 9.
|
16 |
"MMLU": 9.22,
|
|
|
17 |
"WinoGrande": 14.76,
|
18 |
"PiQA": 5.32,
|
19 |
"CommonsenseQA": 9.01,
|
20 |
"Race": 16.19,
|
21 |
"MedMCQA": 1.68,
|
22 |
-
"OpenkookQA": 5.
|
23 |
-
},
|
24 |
-
{
|
25 |
-
"model": "OLMo (1B)",
|
26 |
-
"Average": 8.8,
|
27 |
-
"MMLU": 8.54,
|
28 |
-
"WinoGrande": 6.16,
|
29 |
-
"PiQA": 8.05,
|
30 |
-
"CommonsenseQA": 13.1,
|
31 |
-
"Race": 13.61,
|
32 |
-
"MedMCQA": 2.1,
|
33 |
-
"OpenkookQA": 6.11
|
34 |
-
},
|
35 |
-
{
|
36 |
-
"model": "GPT-Neo (1.3B)",
|
37 |
-
"Average": 7.38,
|
38 |
-
"MMLU": 6.94,
|
39 |
-
"WinoGrande": 10.81,
|
40 |
-
"PiQA": 4.31,
|
41 |
-
"CommonsenseQA": 6.34,
|
42 |
-
"Race": 13.75,
|
43 |
-
"MedMCQA": 2.63,
|
44 |
-
"OpenkookQA": 4.89
|
45 |
-
},
|
46 |
-
{
|
47 |
-
"model": "Cerebras-GPT (1.3B)",
|
48 |
-
"Average": 4.84,
|
49 |
-
"MMLU": 5.37,
|
50 |
-
"WinoGrande": 9.31,
|
51 |
-
"PiQA": 2.16,
|
52 |
-
"CommonsenseQA": 6.2,
|
53 |
-
"Race": 6.9,
|
54 |
-
"MedMCQA": 1.04,
|
55 |
-
"OpenkookQA": 3.46
|
56 |
},
|
57 |
-
|
58 |
"model": "RedPajama (1B)",
|
59 |
-
"Average": 9.
|
60 |
"MMLU": 9.21,
|
|
|
61 |
"WinoGrande": 16.97,
|
62 |
-
"PiQA":
|
63 |
"CommonsenseQA": 11.41,
|
64 |
"Race": 14.35,
|
65 |
"MedMCQA": 1.86,
|
66 |
"OpenkookQA": 3.87
|
67 |
},
|
68 |
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
69 |
"model": "Pythia (1.4B)",
|
70 |
-
"Average": 8.
|
71 |
"MMLU": 9.66,
|
|
|
72 |
"WinoGrande": 11.52,
|
73 |
"PiQA": 4.17,
|
74 |
"CommonsenseQA": 9.01,
|
75 |
"Race": 12.76,
|
76 |
"MedMCQA": 3.19,
|
77 |
-
"OpenkookQA": 5.
|
78 |
},
|
79 |
-
|
80 |
"model": "TinyLLama (1.1B)",
|
81 |
-
"Average": 8.
|
82 |
"MMLU": 8.94,
|
|
|
83 |
"WinoGrande": 12.23,
|
84 |
"PiQA": 3.59,
|
85 |
"CommonsenseQA": 6.06,
|
@@ -87,48 +96,40 @@
|
|
87 |
"MedMCQA": 2.07,
|
88 |
"OpenkookQA": 4.68
|
89 |
},
|
90 |
-
|
91 |
-
"model": "
|
92 |
-
"Average":
|
93 |
-
"MMLU":
|
94 |
-
"
|
95 |
-
"
|
96 |
-
"
|
97 |
-
"
|
98 |
-
"
|
99 |
-
"
|
100 |
-
|
101 |
-
{
|
102 |
-
"model": "Phi-3-mini-128k-instruct (3.8B)",
|
103 |
-
"Average": 39.73,
|
104 |
-
"MMLU": 36.97,
|
105 |
-
"WinoGrande": 46.88,
|
106 |
-
"PiQA": 32.04,
|
107 |
-
"CommonsenseQA": 49.15,
|
108 |
-
"Race": 37.81,
|
109 |
-
"MedMCQA": 22.61,
|
110 |
-
"OpenkookQA": 33.6
|
111 |
},
|
112 |
{
|
113 |
-
"model": "
|
114 |
-
"Average":
|
115 |
-
"MMLU":
|
116 |
-
|
117 |
-
"
|
118 |
-
"
|
119 |
-
"
|
120 |
-
"
|
121 |
-
"
|
|
|
122 |
},
|
123 |
{
|
124 |
-
"model": "
|
125 |
-
"Average":
|
126 |
-
"MMLU":
|
127 |
-
"
|
128 |
-
"
|
129 |
-
"
|
130 |
-
"
|
131 |
-
"
|
132 |
-
"
|
|
|
133 |
}
|
134 |
]
|
|
|
1 |
[
|
2 |
+
{
|
3 |
+
"model": "Phi-3-mini-128k-instruct (3.8B)",
|
4 |
+
"Average": 40.00,
|
5 |
+
"MMLU": 36.97,
|
6 |
+
"ARC":60.94,
|
7 |
+
"WinoGrande": 46.88,
|
8 |
+
"PiQA": 32.04,
|
9 |
+
"CommonsenseQA": 49.15,
|
10 |
+
"Race": 37.81,
|
11 |
+
"MedMCQA": 22.61,
|
12 |
+
"OpenkookQA": 33.60
|
13 |
+
},
|
14 |
+
{
|
15 |
+
"model": "Qwen1.5 (1.8B)",
|
16 |
+
"Average": 21.68,
|
17 |
+
"MMLU": 9.99,
|
18 |
+
"ARC":15.84
|
19 |
+
"WinoGrande": 40.96,
|
20 |
+
"PiQA": 15.52,
|
21 |
+
"CommonsenseQA": 31.13,
|
22 |
+
"Race": 34.91,
|
23 |
+
"MedMCQA": 4.7,
|
24 |
+
"OpenkookQA": 20.37
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"model": "Gemma (2B)",
|
28 |
+
"Average": 16.66,
|
29 |
+
"MMLU": 17.52,
|
30 |
+
"ARC":23.93,
|
31 |
+
"WinoGrande": 16.10,
|
32 |
+
"PiQA": 15.09,
|
33 |
+
"CommonsenseQA": 27.46,
|
34 |
+
"Race": 14.32,
|
35 |
+
"MedMCQA": 4.57,
|
36 |
+
"OpenkookQA": 14.26
|
37 |
},
|
38 |
+
|
39 |
{
|
40 |
+
"model": "SlimPajama-DC (1.3B)",
|
41 |
+
"Average": 9.60,
|
42 |
"MMLU": 9.22,
|
43 |
+
"ARC":14.95,
|
44 |
"WinoGrande": 14.76,
|
45 |
"PiQA": 5.32,
|
46 |
"CommonsenseQA": 9.01,
|
47 |
"Race": 16.19,
|
48 |
"MedMCQA": 1.68,
|
49 |
+
"OpenkookQA": 5.70
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
50 |
},
|
51 |
+
{
|
52 |
"model": "RedPajama (1B)",
|
53 |
+
"Average": 9.00,
|
54 |
"MMLU": 9.21,
|
55 |
+
"ARC":13.5,
|
56 |
"WinoGrande": 16.97,
|
57 |
+
"PiQA": 0.86,
|
58 |
"CommonsenseQA": 11.41,
|
59 |
"Race": 14.35,
|
60 |
"MedMCQA": 1.86,
|
61 |
"OpenkookQA": 3.87
|
62 |
},
|
63 |
{
|
64 |
+
"model": "OLMo (1.2B)",
|
65 |
+
"Average": 8.85,
|
66 |
+
"MMLU": 8.54,
|
67 |
+
"ARC:13.18,
|
68 |
+
"WinoGrande": 6.16,
|
69 |
+
"PiQA": 8.05,
|
70 |
+
"CommonsenseQA": 13.10,
|
71 |
+
"Race": 13.61,
|
72 |
+
"MedMCQA": 2.07,
|
73 |
+
"OpenkookQA": 6.11
|
74 |
+
},
|
75 |
+
{
|
76 |
"model": "Pythia (1.4B)",
|
77 |
+
"Average": 8.79,
|
78 |
"MMLU": 9.66,
|
79 |
+
"ARC":14.69,
|
80 |
"WinoGrande": 11.52,
|
81 |
"PiQA": 4.17,
|
82 |
"CommonsenseQA": 9.01,
|
83 |
"Race": 12.76,
|
84 |
"MedMCQA": 3.19,
|
85 |
+
"OpenkookQA": 5.30
|
86 |
},
|
87 |
+
{
|
88 |
"model": "TinyLLama (1.1B)",
|
89 |
+
"Average": 8.45,
|
90 |
"MMLU": 8.94,
|
91 |
+
"ARC":13.31,
|
92 |
"WinoGrande": 12.23,
|
93 |
"PiQA": 3.59,
|
94 |
"CommonsenseQA": 6.06,
|
|
|
96 |
"MedMCQA": 2.07,
|
97 |
"OpenkookQA": 4.68
|
98 |
},
|
99 |
+
{
|
100 |
+
"model": "OPT (1.3B)",
|
101 |
+
"Average": 7.89,
|
102 |
+
"MMLU": 7.40,
|
103 |
+
"ARC":11.83,
|
104 |
+
"WinoGrande": 12.47,
|
105 |
+
"PiQA": 4.48,
|
106 |
+
"CommonsenseQA": 7.61,
|
107 |
+
"Race": 13.61,
|
108 |
+
"MedMCQA": 1.25,
|
109 |
+
"OpenkookQA": 4.48
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
110 |
},
|
111 |
{
|
112 |
+
"model": "GPT-Neo (1.3B)",
|
113 |
+
"Average": 7.42,
|
114 |
+
"MMLU": 6.94,
|
115 |
+
"ARC": 6.69,
|
116 |
+
"WinoGrande": 10.81,
|
117 |
+
"PiQA": 4.31,
|
118 |
+
"CommonsenseQA": 6.34,
|
119 |
+
"Race": 13.75,
|
120 |
+
"MedMCQA": 2.63,
|
121 |
+
"OpenkookQA": 4.89
|
122 |
},
|
123 |
{
|
124 |
+
"model": "Cerebras-GPT (1.3B)",
|
125 |
+
"Average": 4.86,
|
126 |
+
"MMLU": 5.37,
|
127 |
+
"ARC":4.43,
|
128 |
+
"WinoGrande": 9.31,
|
129 |
+
"PiQA": 2.16,
|
130 |
+
"CommonsenseQA": 6.2,
|
131 |
+
"Race": 6.9,
|
132 |
+
"MedMCQA": 1.04,
|
133 |
+
"OpenkookQA": 3.46
|
134 |
}
|
135 |
]
|