Upload external_comparison.json with huggingface_hub
Browse files- external_comparison.json +437 -0
external_comparison.json
ADDED
|
@@ -0,0 +1,437 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"name": "shared_bpe_8000",
|
| 4 |
+
"source": "ours",
|
| 5 |
+
"algorithm": "bpe",
|
| 6 |
+
"architecture": "shared",
|
| 7 |
+
"vocab_size": 8000,
|
| 8 |
+
"fertility_ar": 1.4755656463463651,
|
| 9 |
+
"fertility_az": 1.7595859759130557,
|
| 10 |
+
"fertility_overall": 1.5744684296670828,
|
| 11 |
+
"disparity": 0.16141315823985886,
|
| 12 |
+
"cpt_ar": 3.812953050264278,
|
| 13 |
+
"cpt_az": 3.0042165735724513,
|
| 14 |
+
"exact_match_ar": 0.9990480507706255,
|
| 15 |
+
"exact_match_az": 0.9992363821483116
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"name": "concat_bpe_8000",
|
| 19 |
+
"source": "ours",
|
| 20 |
+
"algorithm": "bpe",
|
| 21 |
+
"architecture": "concatenated",
|
| 22 |
+
"vocab_size": 8000,
|
| 23 |
+
"fertility_ar": 1.677255929872786,
|
| 24 |
+
"fertility_az": 1.5274486975717831,
|
| 25 |
+
"fertility_overall": 1.625089409755206,
|
| 26 |
+
"disparity": 0.0893168595399542,
|
| 27 |
+
"cpt_ar": 3.3745329363977343,
|
| 28 |
+
"cpt_az": 3.4826508889784487,
|
| 29 |
+
"exact_match_ar": 0.9989120580235721,
|
| 30 |
+
"exact_match_az": 0.9964364500254539
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"name": "shared_unigram_8000",
|
| 34 |
+
"source": "ours",
|
| 35 |
+
"algorithm": "unigram",
|
| 36 |
+
"architecture": "shared",
|
| 37 |
+
"vocab_size": 8000,
|
| 38 |
+
"fertility_ar": 1.5098217608254827,
|
| 39 |
+
"fertility_az": 1.7286510868870875,
|
| 40 |
+
"fertility_overall": 1.5860234519252308,
|
| 41 |
+
"disparity": 0.12658964421540223,
|
| 42 |
+
"cpt_ar": 3.7495598617218993,
|
| 43 |
+
"cpt_az": 3.072876133532698,
|
| 44 |
+
"exact_match_ar": 0.9990480507706255,
|
| 45 |
+
"exact_match_az": 0.9992363821483116
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"name": "concat_unigram_8000",
|
| 49 |
+
"source": "ours",
|
| 50 |
+
"algorithm": "unigram",
|
| 51 |
+
"architecture": "concatenated",
|
| 52 |
+
"vocab_size": 8000,
|
| 53 |
+
"fertility_ar": 1.6828221737842166,
|
| 54 |
+
"fertility_az": 1.5327330595500248,
|
| 55 |
+
"fertility_overall": 1.630557495524919,
|
| 56 |
+
"disparity": 0.08918893307465849,
|
| 57 |
+
"cpt_ar": 3.3890978047599254,
|
| 58 |
+
"cpt_az": 3.4939403000980764,
|
| 59 |
+
"exact_match_ar": 0.9989120580235721,
|
| 60 |
+
"exact_match_az": 0.9964364500254539
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"name": "shared_wordpiece_8000",
|
| 64 |
+
"source": "ours",
|
| 65 |
+
"algorithm": "wordpiece",
|
| 66 |
+
"architecture": "shared",
|
| 67 |
+
"vocab_size": 8000,
|
| 68 |
+
"fertility_ar": 1.4715587661386824,
|
| 69 |
+
"fertility_az": 1.7600376405371791,
|
| 70 |
+
"fertility_overall": 1.5720141231575526,
|
| 71 |
+
"disparity": 0.16390494598197936,
|
| 72 |
+
"cpt_ar": 3.819823570325556,
|
| 73 |
+
"cpt_az": 3.001677181426908,
|
| 74 |
+
"exact_match_ar": 0.0,
|
| 75 |
+
"exact_match_az": 0.0
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"name": "concat_wordpiece_8000",
|
| 79 |
+
"source": "ours",
|
| 80 |
+
"algorithm": "wordpiece",
|
| 81 |
+
"architecture": "concatenated",
|
| 82 |
+
"vocab_size": 8000,
|
| 83 |
+
"fertility_ar": 1.679880351462672,
|
| 84 |
+
"fertility_az": 1.5381307082241507,
|
| 85 |
+
"fertility_overall": 1.6305196797375283,
|
| 86 |
+
"disparity": 0.08438079718897826,
|
| 87 |
+
"cpt_ar": 3.366673383613362,
|
| 88 |
+
"cpt_az": 3.4556741357085015,
|
| 89 |
+
"exact_match_ar": 0.0,
|
| 90 |
+
"exact_match_az": 0.0
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"name": "shared_bbpe_8000",
|
| 94 |
+
"source": "ours",
|
| 95 |
+
"algorithm": "bbpe",
|
| 96 |
+
"architecture": "shared",
|
| 97 |
+
"vocab_size": 8000,
|
| 98 |
+
"fertility_ar": 1.5432842580927282,
|
| 99 |
+
"fertility_az": 2.037850131276259,
|
| 100 |
+
"fertility_overall": 1.7155041180862605,
|
| 101 |
+
"disparity": 0.24269001218151182,
|
| 102 |
+
"cpt_ar": 3.6389791874263038,
|
| 103 |
+
"cpt_az": 2.5723083391604016,
|
| 104 |
+
"exact_match_ar": 0.9994106980961015,
|
| 105 |
+
"exact_match_az": 0.9994909214322077
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"name": "concat_bbpe_8000",
|
| 109 |
+
"source": "ours",
|
| 110 |
+
"algorithm": "bbpe",
|
| 111 |
+
"architecture": "concatenated",
|
| 112 |
+
"vocab_size": 8000,
|
| 113 |
+
"fertility_ar": 1.733122664722859,
|
| 114 |
+
"fertility_az": 1.8376823362685084,
|
| 115 |
+
"fertility_overall": 1.7695328842122233,
|
| 116 |
+
"disparity": 0.056897576627940066,
|
| 117 |
+
"cpt_ar": 3.268439481034545,
|
| 118 |
+
"cpt_az": 2.872055574482044,
|
| 119 |
+
"exact_match_ar": 0.999909338168631,
|
| 120 |
+
"exact_match_az": 0.9997454607161038
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"name": "shared_bpe_16000",
|
| 124 |
+
"source": "ours",
|
| 125 |
+
"algorithm": "bpe",
|
| 126 |
+
"architecture": "shared",
|
| 127 |
+
"vocab_size": 16000,
|
| 128 |
+
"fertility_ar": 1.3322457895485136,
|
| 129 |
+
"fertility_az": 1.545385811386613,
|
| 130 |
+
"fertility_overall": 1.406466326610023,
|
| 131 |
+
"disparity": 0.13792026577936378,
|
| 132 |
+
"cpt_ar": 4.211736441809671,
|
| 133 |
+
"cpt_az": 3.426166367193676,
|
| 134 |
+
"exact_match_ar": 0.9990480507706255,
|
| 135 |
+
"exact_match_az": 0.9992363821483116
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"name": "concat_bpe_16000",
|
| 139 |
+
"source": "ours",
|
| 140 |
+
"algorithm": "bpe",
|
| 141 |
+
"architecture": "concatenated",
|
| 142 |
+
"vocab_size": 16000,
|
| 143 |
+
"fertility_ar": 1.4863282848922221,
|
| 144 |
+
"fertility_az": 1.3611285082800872,
|
| 145 |
+
"fertility_overall": 1.4427306790554728,
|
| 146 |
+
"disparity": 0.08423426902705648,
|
| 147 |
+
"cpt_ar": 3.806560946253488,
|
| 148 |
+
"cpt_az": 3.9116836675170874,
|
| 149 |
+
"exact_match_ar": 0.9989120580235721,
|
| 150 |
+
"exact_match_az": 0.9964364500254539
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"name": "shared_unigram_16000",
|
| 154 |
+
"source": "ours",
|
| 155 |
+
"algorithm": "unigram",
|
| 156 |
+
"architecture": "shared",
|
| 157 |
+
"vocab_size": 16000,
|
| 158 |
+
"fertility_ar": 1.383798509647741,
|
| 159 |
+
"fertility_az": 1.542776425502961,
|
| 160 |
+
"fertility_overall": 1.43915848471923,
|
| 161 |
+
"disparity": 0.10304663282846806,
|
| 162 |
+
"cpt_ar": 4.074326867754011,
|
| 163 |
+
"cpt_az": 3.4475943366681565,
|
| 164 |
+
"exact_match_ar": 0.9990480507706255,
|
| 165 |
+
"exact_match_az": 0.9992363821483116
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"name": "concat_unigram_16000",
|
| 169 |
+
"source": "ours",
|
| 170 |
+
"algorithm": "unigram",
|
| 171 |
+
"architecture": "concatenated",
|
| 172 |
+
"vocab_size": 16000,
|
| 173 |
+
"fertility_ar": 1.5133852390561142,
|
| 174 |
+
"fertility_az": 1.3859550062802124,
|
| 175 |
+
"fertility_overall": 1.4690109341605053,
|
| 176 |
+
"disparity": 0.08420211158883709,
|
| 177 |
+
"cpt_ar": 3.7636569223952008,
|
| 178 |
+
"cpt_az": 3.865353143030446,
|
| 179 |
+
"exact_match_ar": 0.9989120580235721,
|
| 180 |
+
"exact_match_az": 0.9964364500254539
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"name": "shared_wordpiece_16000",
|
| 184 |
+
"source": "ours",
|
| 185 |
+
"algorithm": "wordpiece",
|
| 186 |
+
"architecture": "shared",
|
| 187 |
+
"vocab_size": 16000,
|
| 188 |
+
"fertility_ar": 1.3278822099406427,
|
| 189 |
+
"fertility_az": 1.5399298737530456,
|
| 190 |
+
"fertility_overall": 1.4017223613822598,
|
| 191 |
+
"disparity": 0.1376995585491242,
|
| 192 |
+
"cpt_ar": 4.223103615213092,
|
| 193 |
+
"cpt_az": 3.43676962420351,
|
| 194 |
+
"exact_match_ar": 0.0,
|
| 195 |
+
"exact_match_az": 0.0
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"name": "concat_wordpiece_16000",
|
| 199 |
+
"source": "ours",
|
| 200 |
+
"algorithm": "wordpiece",
|
| 201 |
+
"architecture": "concatenated",
|
| 202 |
+
"vocab_size": 16000,
|
| 203 |
+
"fertility_ar": 1.4835885687430344,
|
| 204 |
+
"fertility_az": 1.3635334584201908,
|
| 205 |
+
"fertility_overall": 1.441782460775622,
|
| 206 |
+
"disparity": 0.08092210526032824,
|
| 207 |
+
"cpt_ar": 3.81082194270718,
|
| 208 |
+
"cpt_az": 3.9034221417934196,
|
| 209 |
+
"exact_match_ar": 0.0,
|
| 210 |
+
"exact_match_az": 0.0
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"name": "shared_bbpe_16000",
|
| 214 |
+
"source": "ours",
|
| 215 |
+
"algorithm": "bbpe",
|
| 216 |
+
"architecture": "shared",
|
| 217 |
+
"vocab_size": 16000,
|
| 218 |
+
"fertility_ar": 1.407745203379972,
|
| 219 |
+
"fertility_az": 1.8543152757754426,
|
| 220 |
+
"fertility_overall": 1.563251758755881,
|
| 221 |
+
"disparity": 0.24082747860054313,
|
| 222 |
+
"cpt_ar": 3.9802348152214964,
|
| 223 |
+
"cpt_az": 2.8299292509695566,
|
| 224 |
+
"exact_match_ar": 0.9994106980961015,
|
| 225 |
+
"exact_match_az": 0.9994909214322077
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"name": "concat_bbpe_16000",
|
| 229 |
+
"source": "ours",
|
| 230 |
+
"algorithm": "bbpe",
|
| 231 |
+
"architecture": "concatenated",
|
| 232 |
+
"vocab_size": 16000,
|
| 233 |
+
"fertility_ar": 1.5508741409303313,
|
| 234 |
+
"fertility_az": 1.7047474509939684,
|
| 235 |
+
"fertility_overall": 1.6044565681716603,
|
| 236 |
+
"disparity": 0.09026164548532975,
|
| 237 |
+
"cpt_ar": 3.653251881231358,
|
| 238 |
+
"cpt_az": 3.0920455646111016,
|
| 239 |
+
"exact_match_ar": 0.999909338168631,
|
| 240 |
+
"exact_match_az": 0.9997454607161038
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"name": "shared_bpe_32000",
|
| 244 |
+
"source": "ours",
|
| 245 |
+
"algorithm": "bpe",
|
| 246 |
+
"architecture": "shared",
|
| 247 |
+
"vocab_size": 32000,
|
| 248 |
+
"fertility_ar": 1.2326026481342502,
|
| 249 |
+
"fertility_az": 1.3678412557544855,
|
| 250 |
+
"fertility_overall": 1.2796960189731112,
|
| 251 |
+
"disparity": 0.09887010429849855,
|
| 252 |
+
"cpt_ar": 4.537455476649384,
|
| 253 |
+
"cpt_az": 3.8681871063030555,
|
| 254 |
+
"exact_match_ar": 0.9990480507706255,
|
| 255 |
+
"exact_match_az": 0.9992363821483116
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"name": "concat_bpe_32000",
|
| 259 |
+
"source": "ours",
|
| 260 |
+
"algorithm": "bpe",
|
| 261 |
+
"architecture": "concatenated",
|
| 262 |
+
"vocab_size": 32000,
|
| 263 |
+
"fertility_ar": 1.3077466165859795,
|
| 264 |
+
"fertility_az": 1.1981161479838143,
|
| 265 |
+
"fertility_overall": 1.2695706222898995,
|
| 266 |
+
"disparity": 0.08383158267185423,
|
| 267 |
+
"cpt_ar": 4.388664556275568,
|
| 268 |
+
"cpt_az": 4.487021365858093,
|
| 269 |
+
"exact_match_ar": 0.0,
|
| 270 |
+
"exact_match_az": 0.0
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"name": "shared_unigram_32000",
|
| 274 |
+
"source": "ours",
|
| 275 |
+
"algorithm": "unigram",
|
| 276 |
+
"architecture": "shared",
|
| 277 |
+
"vocab_size": 32000,
|
| 278 |
+
"fertility_ar": 1.3027084505800648,
|
| 279 |
+
"fertility_az": 1.3926146506382522,
|
| 280 |
+
"fertility_overall": 1.3340159750699838,
|
| 281 |
+
"disparity": 0.06455928064305677,
|
| 282 |
+
"cpt_ar": 4.308287618942264,
|
| 283 |
+
"cpt_az": 3.8153634896854167,
|
| 284 |
+
"exact_match_ar": 0.9990480507706255,
|
| 285 |
+
"exact_match_az": 0.9992363821483116
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"name": "concat_unigram_32000",
|
| 289 |
+
"source": "ours",
|
| 290 |
+
"algorithm": "unigram",
|
| 291 |
+
"architecture": "concatenated",
|
| 292 |
+
"vocab_size": 32000,
|
| 293 |
+
"fertility_ar": 1.392269196943319,
|
| 294 |
+
"fertility_az": 1.2705202940282205,
|
| 295 |
+
"fertility_overall": 1.3498732692189985,
|
| 296 |
+
"disparity": 0.08744638119007023,
|
| 297 |
+
"cpt_ar": 4.071549633983907,
|
| 298 |
+
"cpt_az": 4.204670213900507,
|
| 299 |
+
"exact_match_ar": 0.9989120580235721,
|
| 300 |
+
"exact_match_az": 0.9964364500254539
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"name": "shared_wordpiece_32000",
|
| 304 |
+
"source": "ours",
|
| 305 |
+
"algorithm": "wordpiece",
|
| 306 |
+
"architecture": "shared",
|
| 307 |
+
"vocab_size": 32000,
|
| 308 |
+
"fertility_ar": 1.2274699882356,
|
| 309 |
+
"fertility_az": 1.3618102283002589,
|
| 310 |
+
"fertility_overall": 1.2742505256521945,
|
| 311 |
+
"disparity": 0.0986482824646834,
|
| 312 |
+
"cpt_ar": 4.553239284995183,
|
| 313 |
+
"cpt_az": 3.883101533084343,
|
| 314 |
+
"exact_match_ar": 0.0,
|
| 315 |
+
"exact_match_az": 0.0
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"name": "concat_wordpiece_32000",
|
| 319 |
+
"source": "ours",
|
| 320 |
+
"algorithm": "wordpiece",
|
| 321 |
+
"architecture": "concatenated",
|
| 322 |
+
"vocab_size": 32000,
|
| 323 |
+
"fertility_ar": 1.3416327230816638,
|
| 324 |
+
"fertility_az": 1.2307158917884393,
|
| 325 |
+
"fertility_overall": 1.303008786024938,
|
| 326 |
+
"disparity": 0.08267302174804893,
|
| 327 |
+
"cpt_ar": 4.201360994624003,
|
| 328 |
+
"cpt_az": 4.313461511548837,
|
| 329 |
+
"exact_match_ar": 0.001042611060743427,
|
| 330 |
+
"exact_match_az": 0.00033938571186153065
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"name": "shared_bbpe_32000",
|
| 334 |
+
"source": "ours",
|
| 335 |
+
"algorithm": "bbpe",
|
| 336 |
+
"architecture": "shared",
|
| 337 |
+
"vocab_size": 32000,
|
| 338 |
+
"fertility_ar": 1.317581832517785,
|
| 339 |
+
"fertility_az": 1.7086120162163183,
|
| 340 |
+
"fertility_overall": 1.4537480484685892,
|
| 341 |
+
"disparity": 0.22885838328847793,
|
| 342 |
+
"cpt_ar": 4.240919303648946,
|
| 343 |
+
"cpt_az": 3.0730455542376744,
|
| 344 |
+
"exact_match_ar": 0.9994106980961015,
|
| 345 |
+
"exact_match_az": 0.9994909214322077
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"name": "concat_bbpe_32000",
|
| 349 |
+
"source": "ours",
|
| 350 |
+
"algorithm": "bbpe",
|
| 351 |
+
"architecture": "concatenated",
|
| 352 |
+
"vocab_size": 32000,
|
| 353 |
+
"fertility_ar": 1.3463628334026025,
|
| 354 |
+
"fertility_az": 1.2326871301490916,
|
| 355 |
+
"fertility_overall": 1.306778190060823,
|
| 356 |
+
"disparity": 0.08443170030639024,
|
| 357 |
+
"cpt_ar": 4.190301018162676,
|
| 358 |
+
"cpt_az": 4.3094384241989285,
|
| 359 |
+
"exact_match_ar": 0.9989120580235721,
|
| 360 |
+
"exact_match_az": 0.9964364500254539
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"name": "CaMeLBERT-MSA",
|
| 364 |
+
"source": "external_msa",
|
| 365 |
+
"algorithm": "WordPiece",
|
| 366 |
+
"architecture": "shared",
|
| 367 |
+
"vocab_size": 30000,
|
| 368 |
+
"fertility_ar": 1.816901627794181,
|
| 369 |
+
"fertility_az": 3.173471216645639,
|
| 370 |
+
"fertility_overall": 2.2892921369888652,
|
| 371 |
+
"disparity": 0.42747184273672184,
|
| 372 |
+
"cpt_ar": 3.1140517054288748,
|
| 373 |
+
"cpt_az": 1.6388977943422216,
|
| 374 |
+
"exact_match_ar": 0.2987307343608341,
|
| 375 |
+
"exact_match_az": 0.3885117936534872
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"name": "Asafaya-BERT",
|
| 379 |
+
"source": "external_msa",
|
| 380 |
+
"algorithm": "WordPiece",
|
| 381 |
+
"architecture": "shared",
|
| 382 |
+
"vocab_size": 32000,
|
| 383 |
+
"fertility_ar": 1.7951679730188856,
|
| 384 |
+
"fertility_az": 2.794867411003224,
|
| 385 |
+
"fertility_overall": 2.143287620128837,
|
| 386 |
+
"disparity": 0.35769118565288,
|
| 387 |
+
"cpt_ar": 3.1312619206100933,
|
| 388 |
+
"cpt_az": 1.8618984630783484,
|
| 389 |
+
"exact_match_ar": 0.19895738893925657,
|
| 390 |
+
"exact_match_az": 0.15238418462582726
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"name": "Aranizer-SP-86k",
|
| 394 |
+
"source": "external_msa",
|
| 395 |
+
"algorithm": "SentencePiece",
|
| 396 |
+
"architecture": "shared",
|
| 397 |
+
"vocab_size": 86000,
|
| 398 |
+
"fertility_ar": 1.5940707675919776,
|
| 399 |
+
"fertility_az": 2.5232344233957935,
|
| 400 |
+
"fertility_overall": 1.9176281406140119,
|
| 401 |
+
"disparity": 0.3682430959202508,
|
| 402 |
+
"cpt_ar": 3.532878084819776,
|
| 403 |
+
"cpt_az": 2.06546468780593,
|
| 404 |
+
"exact_match_ar": 0.9976881233000907,
|
| 405 |
+
"exact_match_az": 0.9955031393178347
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"name": "DarijaBERT-ar",
|
| 409 |
+
"source": "external_darija",
|
| 410 |
+
"algorithm": "WordPiece",
|
| 411 |
+
"architecture": "shared",
|
| 412 |
+
"vocab_size": 80000,
|
| 413 |
+
"fertility_ar": 1.4174523931345464,
|
| 414 |
+
"fertility_az": 2.4040608107988306,
|
| 415 |
+
"fertility_overall": 1.7610134287249042,
|
| 416 |
+
"disparity": 0.41039245481334147,
|
| 417 |
+
"cpt_ar": 3.9815436288421893,
|
| 418 |
+
"cpt_az": 2.170223436755666,
|
| 419 |
+
"exact_match_ar": 0.13721668177697188,
|
| 420 |
+
"exact_match_az": 0.07984048871542508
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"name": "DarijaBERT-az",
|
| 424 |
+
"source": "external_darija",
|
| 425 |
+
"algorithm": "WordPiece",
|
| 426 |
+
"architecture": "shared",
|
| 427 |
+
"vocab_size": 110000,
|
| 428 |
+
"fertility_ar": 1.6053190877519825,
|
| 429 |
+
"fertility_az": 1.5171613637119812,
|
| 430 |
+
"fertility_overall": 1.5746204251172413,
|
| 431 |
+
"disparity": 0.05491601309210961,
|
| 432 |
+
"cpt_ar": 3.476952784741954,
|
| 433 |
+
"cpt_az": 3.451281188374697,
|
| 434 |
+
"exact_match_ar": 0.14777878513145964,
|
| 435 |
+
"exact_match_az": 0.07984048871542508
|
| 436 |
+
}
|
| 437 |
+
]
|