Spaces:
Running
Running
Update afrobench lite
#7
by
JessicaOjo
- opened
leaderboard_json/afrobench_lite.json
ADDED
|
@@ -0,0 +1,176 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"NLI": {
|
| 3 |
+
"afrixnli": {
|
| 4 |
+
"Aya-101 13B": 52.4,
|
| 5 |
+
"Gemma1.1 7b": 34.6,
|
| 6 |
+
"LLaMa2 7b": 33.7,
|
| 7 |
+
"LLaMa3 8B": 35.6,
|
| 8 |
+
"LLaMa3.1 8B": 36.4,
|
| 9 |
+
"LLaMAX3 8B": 41.4,
|
| 10 |
+
"Gemma2 9b": 40.9,
|
| 11 |
+
"Gemma2 27b": 43.7,
|
| 12 |
+
"LLaMa3.1 70B": 38.4,
|
| 13 |
+
"AfroLLaMa 8B": 36.3,
|
| 14 |
+
"Gemini 1.5 pro": 63.6,
|
| 15 |
+
"GPT-4o (Aug)": 66.3,
|
| 16 |
+
"Gemma 3 27b": 51.2,
|
| 17 |
+
"Gemini-2.0 Flash": 66.5,
|
| 18 |
+
"GPT-4.1 (April)": 67.5,
|
| 19 |
+
"LLaMa 4 405B": 45.5,
|
| 20 |
+
"Lugha-Llama 8B": 36.7,
|
| 21 |
+
"Gemini-2.5 Flash": 69.9,
|
| 22 |
+
"Claude 4.0 Sonnet": 68.1,
|
| 23 |
+
"Claude 3.7 Sonnet": 59.8,
|
| 24 |
+
"Claude 4.5 Sonnet": 71.7
|
| 25 |
+
}
|
| 26 |
+
},
|
| 27 |
+
"Intent": {
|
| 28 |
+
"injongointent": {
|
| 29 |
+
"AfroLLaMa 8B": 0.9,
|
| 30 |
+
"LLaMAX3 8B": 6.3,
|
| 31 |
+
"LLaMa2 7b": 1.4,
|
| 32 |
+
"LLaMa3.1 70B": 36.9,
|
| 33 |
+
"LLaMa3.1 8B": 6.3,
|
| 34 |
+
"LLaMa3 8B": 2.3,
|
| 35 |
+
"Gemma1.1 7b": 10.2,
|
| 36 |
+
"Gemma2 27b": 37.8,
|
| 37 |
+
"Gemma2 9b": 34.1,
|
| 38 |
+
"Aya-101 13B": 48.2,
|
| 39 |
+
"Gemini 1.5 pro": 76.3,
|
| 40 |
+
"GPT-4o (Aug)": 78.3,
|
| 41 |
+
"Gemma 3 27b": 55.2,
|
| 42 |
+
"Gemini-2.0 Flash": 73.9,
|
| 43 |
+
"GPT-4.1 (April)": 84.4,
|
| 44 |
+
"LLaMa 4 405B": 73.9,
|
| 45 |
+
"Lugha-Llama 8B": 4.1,
|
| 46 |
+
"Gemini-2.5 Flash": 87.9,
|
| 47 |
+
"Claude 4.0 Sonnet": 80.4,
|
| 48 |
+
"Claude 3.7 Sonnet": 73.4,
|
| 49 |
+
"Claude 4.5 Sonnet": 81.6
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"MT(en/fr-xx)": {
|
| 53 |
+
"flores - en_xx": {
|
| 54 |
+
"AfroLLaMa 8B": 10.7,
|
| 55 |
+
"LLaMAX3 8B": 27.7,
|
| 56 |
+
"LLaMa2 7b": 9.2,
|
| 57 |
+
"LLaMa3.1 70B": 27.4,
|
| 58 |
+
"LLaMa3.1 8B": 14.5,
|
| 59 |
+
"LLaMa3 8B": 11.8,
|
| 60 |
+
"Aya-101 13B": 31.1,
|
| 61 |
+
"Gemma1.1 7b": 11.7,
|
| 62 |
+
"Gemma2 27b": 38.4,
|
| 63 |
+
"Gemma2 9b": 36.5,
|
| 64 |
+
"Gemini 1.5 pro": 43.0,
|
| 65 |
+
"GPT-4o (Aug)": 42.4,
|
| 66 |
+
"Gemma 3 27b": 33.1,
|
| 67 |
+
"Gemini-2.0 Flash": 59.1,
|
| 68 |
+
"GPT-4.1 (April)": 47.3,
|
| 69 |
+
"LLaMa 4 405B": 42.8,
|
| 70 |
+
"Lugha-Llama 8B": 22.1,
|
| 71 |
+
"Gemini-2.5 Flash": 46.5,
|
| 72 |
+
"Claude 4.0 Sonnet": 46.0,
|
| 73 |
+
"Claude 3.7 Sonnet": 44.0
|
| 74 |
+
}
|
| 75 |
+
},
|
| 76 |
+
"MMLU": {
|
| 77 |
+
"afrimmlu": {
|
| 78 |
+
"AfroLLaMa 8B": 25.8,
|
| 79 |
+
"Aya-101 13B": 30.2,
|
| 80 |
+
"Gemma1.1 7b": 27.3,
|
| 81 |
+
"LLaMa2 7b": 25.9,
|
| 82 |
+
"LLaMa3 8B": 28.4,
|
| 83 |
+
"LLaMAX3 8B": 29.6,
|
| 84 |
+
"Gemma2 9b": 36.1,
|
| 85 |
+
"Gemma2 27b": 40.8,
|
| 86 |
+
"LLaMa3.1 8B": 31.6,
|
| 87 |
+
"LLaMa3.1 70B": 40.6,
|
| 88 |
+
"GPT-4o (Aug)": 63.1,
|
| 89 |
+
"Gemini 1.5 pro": 62.6,
|
| 90 |
+
"Gemma 3 27b": 44.4,
|
| 91 |
+
"Gemini-2.0 Flash": 57.8,
|
| 92 |
+
"GPT-4.1 (April)": 60.2,
|
| 93 |
+
"LLaMa 4 405B": 15.8,
|
| 94 |
+
"Lugha-Llama 8B": 25.2,
|
| 95 |
+
"Gemini-2.5 Flash": 67.7,
|
| 96 |
+
"Claude 4.0 Sonnet": 75.5,
|
| 97 |
+
"Claude 3.7 Sonnet": 66.7,
|
| 98 |
+
"Claude 4.5 Sonnet": 58.6
|
| 99 |
+
}
|
| 100 |
+
},
|
| 101 |
+
"Math": {
|
| 102 |
+
"afrimgsm": {
|
| 103 |
+
"AfroLLaMa 8B": 0.3,
|
| 104 |
+
"Aya-101 13B": 4.9,
|
| 105 |
+
"Gemma1.1 7b": 5.0,
|
| 106 |
+
"LLaMa2 7b": 1.9,
|
| 107 |
+
"LLaMa3 8B": 5.7,
|
| 108 |
+
"LLaMa3.1 8B": 7.8,
|
| 109 |
+
"LLaMAX3 8B": 5.2,
|
| 110 |
+
"Gemma2 9b": 21.7,
|
| 111 |
+
"Gemma2 27b": 31.1,
|
| 112 |
+
"LLaMa3.1 70B": 26.5,
|
| 113 |
+
"GPT-4o (Aug)": 57.3,
|
| 114 |
+
"Gemini 1.5 pro": 57.7,
|
| 115 |
+
"Gemma 3 27b": 47.5,
|
| 116 |
+
"Gemini-2.0 Flash": 67.5,
|
| 117 |
+
"GPT-4.1 (April)": 59.5,
|
| 118 |
+
"LLaMa 4 405B": 65.0,
|
| 119 |
+
"Lugha-Llama 8B": 1.8,
|
| 120 |
+
"Gemini-2.5 Flash": 70.6,
|
| 121 |
+
"Claude 4.0 Sonnet": 66.9,
|
| 122 |
+
"Claude 3.7 Sonnet": 35.2,
|
| 123 |
+
"Claude 4.5 Sonnet": 73.1
|
| 124 |
+
}
|
| 125 |
+
},
|
| 126 |
+
"Topic": {
|
| 127 |
+
"sib": {
|
| 128 |
+
"AfroLLaMa 8B": 13.9,
|
| 129 |
+
"LLaMAX3 8B": 62.6,
|
| 130 |
+
"LLaMa2 7b": 18.0,
|
| 131 |
+
"LLaMa3.1 70B": 64.3,
|
| 132 |
+
"LLaMa3.1 8B": 48.0,
|
| 133 |
+
"LLaMa3 8B": 40.1,
|
| 134 |
+
"Aya-101 13B": 76.6,
|
| 135 |
+
"Gemma1.1 7b": 42.7,
|
| 136 |
+
"Gemma2 27b": 68.8,
|
| 137 |
+
"Gemma2 9b": 65.7,
|
| 138 |
+
"Gemini 1.5 pro": 81.3,
|
| 139 |
+
"GPT-4o (Aug)": 82.8,
|
| 140 |
+
"Gemma 3 27b": 74.2,
|
| 141 |
+
"Gemini-2.0 Flash": 84.9,
|
| 142 |
+
"GPT-4.1 (April)": 84.8,
|
| 143 |
+
"LLaMa 4 405B": 80.6,
|
| 144 |
+
"Lugha-Llama 8B": 34.1,
|
| 145 |
+
"Gemini-2.5 Flash": 87.2,
|
| 146 |
+
"Claude 4.0 Sonnet": 83.2,
|
| 147 |
+
"Claude 3.7 Sonnet": 84.9,
|
| 148 |
+
"Claude 4.5 Sonnet": 84.2
|
| 149 |
+
}
|
| 150 |
+
},
|
| 151 |
+
"RC": {
|
| 152 |
+
"belebele": {
|
| 153 |
+
"AfroLLaMa 8B": 24.7,
|
| 154 |
+
"LLaMAX3 8B": 28.9,
|
| 155 |
+
"LLaMa2 7b": 23.5,
|
| 156 |
+
"LLaMa3.1 8B": 36.2,
|
| 157 |
+
"LLaMa3.1 70B": 45.9,
|
| 158 |
+
"LLaMa3 8B": 26.2,
|
| 159 |
+
"Aya-101 13B": 60.3,
|
| 160 |
+
"Gemma1.1 7b": 34.3,
|
| 161 |
+
"Gemma2 27b": 53.5,
|
| 162 |
+
"Gemma2 9b": 50.5,
|
| 163 |
+
"Gemini 1.5 pro": 55.0,
|
| 164 |
+
"GPT-4o (Aug)": 70.4,
|
| 165 |
+
"Gemma 3 27b": 62.4,
|
| 166 |
+
"Gemini-2.0 Flash": 56.8,
|
| 167 |
+
"GPT-4.1 (April)": 64.8,
|
| 168 |
+
"LLaMa 4 405B": 24.6,
|
| 169 |
+
"Lugha-Llama 8B": 23.0,
|
| 170 |
+
"Gemini-2.5 Flash": 42.2,
|
| 171 |
+
"Claude 4.0 Sonnet": 76.2,
|
| 172 |
+
"Claude 3.7 Sonnet": 65.1,
|
| 173 |
+
"Claude 4.5 Sonnet": 74.8
|
| 174 |
+
}
|
| 175 |
+
}
|
| 176 |
+
}
|