Ouaill commited on
Commit
765c3cc
·
verified ·
1 Parent(s): 6368cdb

Upload results/external_datasets_eval.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. results/external_datasets_eval.json +254 -0
results/external_datasets_eval.json ADDED
@@ -0,0 +1,254 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "dataset": "DODa",
4
+ "tokenizer": "DarijaBERT-ar",
5
+ "vocab_size": 80000,
6
+ "source": "external",
7
+ "fertility_ar": 0.0,
8
+ "fertility_az": 2.648,
9
+ "fertility_overall": 2.648,
10
+ "cpt_ar": 0.0,
11
+ "cpt_az": 2.194,
12
+ "cpt_overall": 2.194,
13
+ "gain_pct": 0.0,
14
+ "n_texts": 10000
15
+ },
16
+ {
17
+ "dataset": "DODa",
18
+ "tokenizer": "DarijaBERT-az",
19
+ "vocab_size": 110000,
20
+ "source": "external",
21
+ "fertility_ar": 0.0,
22
+ "fertility_az": 1.246,
23
+ "fertility_overall": 1.246,
24
+ "cpt_ar": 0.0,
25
+ "cpt_az": 4.677,
26
+ "cpt_overall": 4.677,
27
+ "gain_pct": 0.0,
28
+ "n_texts": 10000
29
+ },
30
+ {
31
+ "dataset": "DODa",
32
+ "tokenizer": "DarijaBERT-mix",
33
+ "vocab_size": 160000,
34
+ "source": "external",
35
+ "fertility_ar": 0.0,
36
+ "fertility_az": 1.318,
37
+ "fertility_overall": 1.318,
38
+ "cpt_ar": 0.0,
39
+ "cpt_az": 4.441,
40
+ "cpt_overall": 4.441,
41
+ "gain_pct": 0.0,
42
+ "n_texts": 10000
43
+ },
44
+ {
45
+ "dataset": "DODa",
46
+ "tokenizer": "Ours (80K WP)",
47
+ "vocab_size": 80000,
48
+ "source": "ours",
49
+ "fertility_ar": 0.0,
50
+ "fertility_az": 1.72,
51
+ "fertility_overall": 1.72,
52
+ "cpt_ar": 0.0,
53
+ "cpt_az": 3.467,
54
+ "cpt_overall": 3.467,
55
+ "gain_pct": 35.0,
56
+ "n_texts": 10000
57
+ },
58
+ {
59
+ "dataset": "DODa",
60
+ "tokenizer": "Ours (110K WP)",
61
+ "vocab_size": 110000,
62
+ "source": "ours",
63
+ "fertility_ar": 0.0,
64
+ "fertility_az": 1.68,
65
+ "fertility_overall": 1.68,
66
+ "cpt_ar": 0.0,
67
+ "cpt_az": 3.551,
68
+ "cpt_overall": 3.551,
69
+ "gain_pct": -34.8,
70
+ "n_texts": 10000
71
+ },
72
+ {
73
+ "dataset": "DODa",
74
+ "tokenizer": "Ours (32K BPE)",
75
+ "vocab_size": 32000,
76
+ "source": "ours",
77
+ "fertility_ar": 0.0,
78
+ "fertility_az": 1.901,
79
+ "fertility_overall": 1.901,
80
+ "cpt_ar": 0.0,
81
+ "cpt_az": 3.145,
82
+ "cpt_overall": 3.145,
83
+ "gain_pct": -44.2,
84
+ "n_texts": 10000
85
+ },
86
+ {
87
+ "dataset": "Darija-Wiki",
88
+ "tokenizer": "DarijaBERT-ar",
89
+ "vocab_size": 80000,
90
+ "source": "external",
91
+ "fertility_ar": 1.56,
92
+ "fertility_az": 2.502,
93
+ "fertility_overall": 1.562,
94
+ "cpt_ar": 3.382,
95
+ "cpt_az": 2.589,
96
+ "cpt_overall": 3.381,
97
+ "gain_pct": 0.0,
98
+ "n_texts": 10000
99
+ },
100
+ {
101
+ "dataset": "Darija-Wiki",
102
+ "tokenizer": "DarijaBERT-az",
103
+ "vocab_size": 110000,
104
+ "source": "external",
105
+ "fertility_ar": 1.858,
106
+ "fertility_az": 2.219,
107
+ "fertility_overall": 1.858,
108
+ "cpt_ar": 2.829,
109
+ "cpt_az": 2.946,
110
+ "cpt_overall": 2.829,
111
+ "gain_pct": 0.0,
112
+ "n_texts": 10000
113
+ },
114
+ {
115
+ "dataset": "Darija-Wiki",
116
+ "tokenizer": "DarijaBERT-mix",
117
+ "vocab_size": 160000,
118
+ "source": "external",
119
+ "fertility_ar": 1.502,
120
+ "fertility_az": 2.142,
121
+ "fertility_overall": 1.503,
122
+ "cpt_ar": 3.512,
123
+ "cpt_az": 3.04,
124
+ "cpt_overall": 3.511,
125
+ "gain_pct": 0.0,
126
+ "n_texts": 10000
127
+ },
128
+ {
129
+ "dataset": "Darija-Wiki",
130
+ "tokenizer": "Ours (80K WP)",
131
+ "vocab_size": 80000,
132
+ "source": "ours",
133
+ "fertility_ar": 1.839,
134
+ "fertility_az": 3.265,
135
+ "fertility_overall": 1.841,
136
+ "cpt_ar": 2.883,
137
+ "cpt_az": 2.07,
138
+ "cpt_overall": 2.882,
139
+ "gain_pct": -17.9,
140
+ "n_texts": 10000
141
+ },
142
+ {
143
+ "dataset": "Darija-Wiki",
144
+ "tokenizer": "Ours (110K WP)",
145
+ "vocab_size": 110000,
146
+ "source": "ours",
147
+ "fertility_ar": 1.797,
148
+ "fertility_az": 3.184,
149
+ "fertility_overall": 1.799,
150
+ "cpt_ar": 2.951,
151
+ "cpt_az": 2.124,
152
+ "cpt_overall": 2.95,
153
+ "gain_pct": 3.2,
154
+ "n_texts": 10000
155
+ },
156
+ {
157
+ "dataset": "Darija-Wiki",
158
+ "tokenizer": "Ours (32K BPE)",
159
+ "vocab_size": 32000,
160
+ "source": "ours",
161
+ "fertility_ar": 2.102,
162
+ "fertility_az": 3.805,
163
+ "fertility_overall": 2.104,
164
+ "cpt_ar": 2.519,
165
+ "cpt_az": 1.743,
166
+ "cpt_overall": 2.518,
167
+ "gain_pct": -40.0,
168
+ "n_texts": 10000
169
+ },
170
+ {
171
+ "dataset": "Atlaset",
172
+ "tokenizer": "DarijaBERT-ar",
173
+ "vocab_size": 80000,
174
+ "source": "external",
175
+ "fertility_ar": 1.407,
176
+ "fertility_az": 2.056,
177
+ "fertility_overall": 1.41,
178
+ "cpt_ar": 4.051,
179
+ "cpt_az": 1.803,
180
+ "cpt_overall": 4.041,
181
+ "gain_pct": 0.0,
182
+ "n_texts": 10000
183
+ },
184
+ {
185
+ "dataset": "Atlaset",
186
+ "tokenizer": "DarijaBERT-az",
187
+ "vocab_size": 110000,
188
+ "source": "external",
189
+ "fertility_ar": 1.772,
190
+ "fertility_az": 1.918,
191
+ "fertility_overall": 1.773,
192
+ "cpt_ar": 3.22,
193
+ "cpt_az": 1.965,
194
+ "cpt_overall": 3.215,
195
+ "gain_pct": 0.0,
196
+ "n_texts": 10000
197
+ },
198
+ {
199
+ "dataset": "Atlaset",
200
+ "tokenizer": "DarijaBERT-mix",
201
+ "vocab_size": 160000,
202
+ "source": "external",
203
+ "fertility_ar": 1.379,
204
+ "fertility_az": 1.922,
205
+ "fertility_overall": 1.381,
206
+ "cpt_ar": 4.131,
207
+ "cpt_az": 1.963,
208
+ "cpt_overall": 4.122,
209
+ "gain_pct": 0.0,
210
+ "n_texts": 10000
211
+ },
212
+ {
213
+ "dataset": "Atlaset",
214
+ "tokenizer": "Ours (80K WP)",
215
+ "vocab_size": 80000,
216
+ "source": "ours",
217
+ "fertility_ar": 1.618,
218
+ "fertility_az": 2.293,
219
+ "fertility_overall": 1.621,
220
+ "cpt_ar": 3.594,
221
+ "cpt_az": 1.771,
222
+ "cpt_overall": 3.586,
223
+ "gain_pct": -15.0,
224
+ "n_texts": 10000
225
+ },
226
+ {
227
+ "dataset": "Atlaset",
228
+ "tokenizer": "Ours (110K WP)",
229
+ "vocab_size": 110000,
230
+ "source": "ours",
231
+ "fertility_ar": 1.566,
232
+ "fertility_az": 2.183,
233
+ "fertility_overall": 1.569,
234
+ "cpt_ar": 3.714,
235
+ "cpt_az": 1.846,
236
+ "cpt_overall": 3.706,
237
+ "gain_pct": 11.5,
238
+ "n_texts": 10000
239
+ },
240
+ {
241
+ "dataset": "Atlaset",
242
+ "tokenizer": "Ours (32K BPE)",
243
+ "vocab_size": 32000,
244
+ "source": "ours",
245
+ "fertility_ar": 1.864,
246
+ "fertility_az": 2.571,
247
+ "fertility_overall": 1.867,
248
+ "cpt_ar": 3.119,
249
+ "cpt_az": 1.492,
250
+ "cpt_overall": 3.112,
251
+ "gain_pct": -35.2,
252
+ "n_texts": 10000
253
+ }
254
+ ]