Upload evaluation_report.json with huggingface_hub
Browse files- evaluation_report.json +237 -0
evaluation_report.json
ADDED
|
@@ -0,0 +1,237 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"tokenizer_config": {
|
| 3 |
+
"name": "sudanizer-72k",
|
| 4 |
+
"version": "1.0.0",
|
| 5 |
+
"vocab_size": 72000,
|
| 6 |
+
"min_frequency": 3,
|
| 7 |
+
"limit_alphabet": 5000,
|
| 8 |
+
"pre_tokenizer": "whitespace",
|
| 9 |
+
"dialect_normalization": false,
|
| 10 |
+
"hamza_normalization": true
|
| 11 |
+
},
|
| 12 |
+
"training_metrics": {
|
| 13 |
+
"total_samples": 500,
|
| 14 |
+
"total_characters": 86872,
|
| 15 |
+
"total_tokens": 19153,
|
| 16 |
+
"total_words": 16772,
|
| 17 |
+
"avg_chars_per_token": 4.536,
|
| 18 |
+
"compression_ratio": 0.2205,
|
| 19 |
+
"unk_count": 0,
|
| 20 |
+
"unk_rate_percent": 0.0,
|
| 21 |
+
"avg_subwords_per_word": 1.159,
|
| 22 |
+
"vocab_size": 72000,
|
| 23 |
+
"max_tokens_in_sample": 868,
|
| 24 |
+
"min_tokens_in_sample": 2,
|
| 25 |
+
"avg_tokens_per_sample": 38.31
|
| 26 |
+
},
|
| 27 |
+
"comprehensive_evaluation": {
|
| 28 |
+
"by_category": {
|
| 29 |
+
"pure_dialect": {
|
| 30 |
+
"total_samples": 10,
|
| 31 |
+
"total_characters": 311,
|
| 32 |
+
"total_tokens": 66,
|
| 33 |
+
"total_words": 62,
|
| 34 |
+
"avg_chars_per_token": 4.712,
|
| 35 |
+
"compression_ratio": 0.2122,
|
| 36 |
+
"unk_count": 0,
|
| 37 |
+
"unk_rate_percent": 0.0,
|
| 38 |
+
"avg_subwords_per_word": 1.064,
|
| 39 |
+
"vocab_size": 72000,
|
| 40 |
+
"max_tokens_in_sample": 8,
|
| 41 |
+
"min_tokens_in_sample": 5,
|
| 42 |
+
"avg_tokens_per_sample": 6.6
|
| 43 |
+
},
|
| 44 |
+
"code_switching": {
|
| 45 |
+
"total_samples": 8,
|
| 46 |
+
"total_characters": 349,
|
| 47 |
+
"total_tokens": 156,
|
| 48 |
+
"total_words": 53,
|
| 49 |
+
"avg_chars_per_token": 2.237,
|
| 50 |
+
"compression_ratio": 0.447,
|
| 51 |
+
"unk_count": 0,
|
| 52 |
+
"unk_rate_percent": 0.0,
|
| 53 |
+
"avg_subwords_per_word": 2.984,
|
| 54 |
+
"vocab_size": 72000,
|
| 55 |
+
"max_tokens_in_sample": 27,
|
| 56 |
+
"min_tokens_in_sample": 13,
|
| 57 |
+
"avg_tokens_per_sample": 19.5
|
| 58 |
+
},
|
| 59 |
+
"social_media": {
|
| 60 |
+
"total_samples": 10,
|
| 61 |
+
"total_characters": 378,
|
| 62 |
+
"total_tokens": 129,
|
| 63 |
+
"total_words": 72,
|
| 64 |
+
"avg_chars_per_token": 2.93,
|
| 65 |
+
"compression_ratio": 0.3413,
|
| 66 |
+
"unk_count": 29,
|
| 67 |
+
"unk_rate_percent": 22.4806,
|
| 68 |
+
"avg_subwords_per_word": 1.853,
|
| 69 |
+
"vocab_size": 72000,
|
| 70 |
+
"max_tokens_in_sample": 25,
|
| 71 |
+
"min_tokens_in_sample": 10,
|
| 72 |
+
"avg_tokens_per_sample": 12.9
|
| 73 |
+
},
|
| 74 |
+
"formal": {
|
| 75 |
+
"total_samples": 5,
|
| 76 |
+
"total_characters": 356,
|
| 77 |
+
"total_tokens": 68,
|
| 78 |
+
"total_words": 51,
|
| 79 |
+
"avg_chars_per_token": 5.235,
|
| 80 |
+
"compression_ratio": 0.191,
|
| 81 |
+
"unk_count": 1,
|
| 82 |
+
"unk_rate_percent": 1.4706,
|
| 83 |
+
"avg_subwords_per_word": 1.315,
|
| 84 |
+
"vocab_size": 72000,
|
| 85 |
+
"max_tokens_in_sample": 19,
|
| 86 |
+
"min_tokens_in_sample": 9,
|
| 87 |
+
"avg_tokens_per_sample": 13.6
|
| 88 |
+
},
|
| 89 |
+
"edge_case": {
|
| 90 |
+
"total_samples": 8,
|
| 91 |
+
"total_characters": 355,
|
| 92 |
+
"total_tokens": 76,
|
| 93 |
+
"total_words": 50,
|
| 94 |
+
"avg_chars_per_token": 4.671,
|
| 95 |
+
"compression_ratio": 0.2141,
|
| 96 |
+
"unk_count": 1,
|
| 97 |
+
"unk_rate_percent": 1.3158,
|
| 98 |
+
"avg_subwords_per_word": 1.525,
|
| 99 |
+
"vocab_size": 72000,
|
| 100 |
+
"max_tokens_in_sample": 13,
|
| 101 |
+
"min_tokens_in_sample": 8,
|
| 102 |
+
"avg_tokens_per_sample": 9.5
|
| 103 |
+
},
|
| 104 |
+
"morphology": {
|
| 105 |
+
"total_samples": 8,
|
| 106 |
+
"total_characters": 216,
|
| 107 |
+
"total_tokens": 48,
|
| 108 |
+
"total_words": 40,
|
| 109 |
+
"avg_chars_per_token": 4.5,
|
| 110 |
+
"compression_ratio": 0.2222,
|
| 111 |
+
"unk_count": 0,
|
| 112 |
+
"unk_rate_percent": 0.0,
|
| 113 |
+
"avg_subwords_per_word": 1.205,
|
| 114 |
+
"vocab_size": 72000,
|
| 115 |
+
"max_tokens_in_sample": 8,
|
| 116 |
+
"min_tokens_in_sample": 4,
|
| 117 |
+
"avg_tokens_per_sample": 6.0
|
| 118 |
+
},
|
| 119 |
+
"cultural": {
|
| 120 |
+
"total_samples": 8,
|
| 121 |
+
"total_characters": 267,
|
| 122 |
+
"total_tokens": 59,
|
| 123 |
+
"total_words": 49,
|
| 124 |
+
"avg_chars_per_token": 4.525,
|
| 125 |
+
"compression_ratio": 0.221,
|
| 126 |
+
"unk_count": 0,
|
| 127 |
+
"unk_rate_percent": 0.0,
|
| 128 |
+
"avg_subwords_per_word": 1.187,
|
| 129 |
+
"vocab_size": 72000,
|
| 130 |
+
"max_tokens_in_sample": 12,
|
| 131 |
+
"min_tokens_in_sample": 5,
|
| 132 |
+
"avg_tokens_per_sample": 7.38
|
| 133 |
+
},
|
| 134 |
+
"mixed_content": {
|
| 135 |
+
"total_samples": 8,
|
| 136 |
+
"total_characters": 224,
|
| 137 |
+
"total_tokens": 81,
|
| 138 |
+
"total_words": 33,
|
| 139 |
+
"avg_chars_per_token": 2.765,
|
| 140 |
+
"compression_ratio": 0.3616,
|
| 141 |
+
"unk_count": 1,
|
| 142 |
+
"unk_rate_percent": 1.2346,
|
| 143 |
+
"avg_subwords_per_word": 3.017,
|
| 144 |
+
"vocab_size": 72000,
|
| 145 |
+
"max_tokens_in_sample": 18,
|
| 146 |
+
"min_tokens_in_sample": 5,
|
| 147 |
+
"avg_tokens_per_sample": 10.12
|
| 148 |
+
}
|
| 149 |
+
},
|
| 150 |
+
"by_difficulty": {
|
| 151 |
+
"easy": {
|
| 152 |
+
"total_samples": 4,
|
| 153 |
+
"total_characters": 157,
|
| 154 |
+
"total_tokens": 28,
|
| 155 |
+
"total_words": 28,
|
| 156 |
+
"avg_chars_per_token": 5.607,
|
| 157 |
+
"compression_ratio": 0.1783,
|
| 158 |
+
"unk_count": 0,
|
| 159 |
+
"unk_rate_percent": 0.0,
|
| 160 |
+
"avg_subwords_per_word": 1.0,
|
| 161 |
+
"vocab_size": 72000,
|
| 162 |
+
"max_tokens_in_sample": 9,
|
| 163 |
+
"min_tokens_in_sample": 5,
|
| 164 |
+
"avg_tokens_per_sample": 7.0
|
| 165 |
+
},
|
| 166 |
+
"medium": {
|
| 167 |
+
"total_samples": 22,
|
| 168 |
+
"total_characters": 721,
|
| 169 |
+
"total_tokens": 167,
|
| 170 |
+
"total_words": 129,
|
| 171 |
+
"avg_chars_per_token": 4.317,
|
| 172 |
+
"compression_ratio": 0.2316,
|
| 173 |
+
"unk_count": 4,
|
| 174 |
+
"unk_rate_percent": 2.3952,
|
| 175 |
+
"avg_subwords_per_word": 1.349,
|
| 176 |
+
"vocab_size": 72000,
|
| 177 |
+
"max_tokens_in_sample": 16,
|
| 178 |
+
"min_tokens_in_sample": 4,
|
| 179 |
+
"avg_tokens_per_sample": 7.59
|
| 180 |
+
},
|
| 181 |
+
"hard": {
|
| 182 |
+
"total_samples": 27,
|
| 183 |
+
"total_characters": 1073,
|
| 184 |
+
"total_tokens": 315,
|
| 185 |
+
"total_words": 180,
|
| 186 |
+
"avg_chars_per_token": 3.406,
|
| 187 |
+
"compression_ratio": 0.2936,
|
| 188 |
+
"unk_count": 21,
|
| 189 |
+
"unk_rate_percent": 6.6667,
|
| 190 |
+
"avg_subwords_per_word": 1.794,
|
| 191 |
+
"vocab_size": 72000,
|
| 192 |
+
"max_tokens_in_sample": 27,
|
| 193 |
+
"min_tokens_in_sample": 5,
|
| 194 |
+
"avg_tokens_per_sample": 11.67
|
| 195 |
+
},
|
| 196 |
+
"extreme": {
|
| 197 |
+
"total_samples": 12,
|
| 198 |
+
"total_characters": 505,
|
| 199 |
+
"total_tokens": 173,
|
| 200 |
+
"total_words": 73,
|
| 201 |
+
"avg_chars_per_token": 2.919,
|
| 202 |
+
"compression_ratio": 0.3426,
|
| 203 |
+
"unk_count": 7,
|
| 204 |
+
"unk_rate_percent": 4.0462,
|
| 205 |
+
"avg_subwords_per_word": 2.748,
|
| 206 |
+
"vocab_size": 72000,
|
| 207 |
+
"max_tokens_in_sample": 27,
|
| 208 |
+
"min_tokens_in_sample": 6,
|
| 209 |
+
"avg_tokens_per_sample": 14.42
|
| 210 |
+
}
|
| 211 |
+
},
|
| 212 |
+
"overall": {
|
| 213 |
+
"total_samples": 65,
|
| 214 |
+
"total_characters": 2456,
|
| 215 |
+
"total_tokens": 683,
|
| 216 |
+
"total_words": 410,
|
| 217 |
+
"avg_chars_per_token": 3.596,
|
| 218 |
+
"compression_ratio": 0.2781,
|
| 219 |
+
"unk_count": 32,
|
| 220 |
+
"unk_rate_percent": 4.6852,
|
| 221 |
+
"avg_subwords_per_word": 1.771,
|
| 222 |
+
"vocab_size": 72000,
|
| 223 |
+
"max_tokens_in_sample": 27,
|
| 224 |
+
"min_tokens_in_sample": 4,
|
| 225 |
+
"avg_tokens_per_sample": 10.51
|
| 226 |
+
},
|
| 227 |
+
"problematic_samples_count": 18
|
| 228 |
+
},
|
| 229 |
+
"benchmark_results": {
|
| 230 |
+
"Sudanizer-72K-v2": {
|
| 231 |
+
"avg_chars_per_token": 3.596,
|
| 232 |
+
"unk_rate_percent": 4.6852,
|
| 233 |
+
"avg_subwords_per_word": 1.771,
|
| 234 |
+
"total_tokens": 683
|
| 235 |
+
}
|
| 236 |
+
}
|
| 237 |
+
}
|