@mailwoman/neural-weights-fr-fr 6.0.0 → 6.2.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/country-surface-lexicon-v1.json +326 -0
- package/model-card.json +11 -8
- package/model.onnx +0 -0
- package/package.json +2 -1
- package/tokenizer.model +0 -0
|
@@ -0,0 +1,326 @@
|
|
|
1
|
+
{
|
|
2
|
+
"version": 1,
|
|
3
|
+
"generated_by": "codex/tools/build-country-surface-lexicon.ts (source: @mailwoman/codex COUNTRY_SURFACE_FORMS + ISO2_TO_NAME)",
|
|
4
|
+
"feature_dim": 2,
|
|
5
|
+
"slots": ["country_surface", "country_ambiguous"],
|
|
6
|
+
"bits": {
|
|
7
|
+
"country_surface": 1,
|
|
8
|
+
"country_ambiguous": 2
|
|
9
|
+
},
|
|
10
|
+
"max_ngram": 7,
|
|
11
|
+
"rules": {
|
|
12
|
+
"word_norm": "per whitespace-word: strip leading/trailing chars that are not Unicode letters/digits (keep internal: 'timor-leste', 'u.s.a'); rejoin single-spaced. Applied to BOTH entry keys and scanned tokens.",
|
|
13
|
+
"entries": "case-insensitive; key = word_norm lowercased. country_surface always set; country_ambiguous set for single-token homographs (US region) or curated common-word names.",
|
|
14
|
+
"code_entries": "case-SENSITIVE exact: word_norm(token) == key (keys uppercase; 'in' the word ≠ 'IN' India). n-gram length 1. country_ambiguous set when the code is also a US-state abbreviation.",
|
|
15
|
+
"scan": "longest-first n-gram over whitespace words, left to right, non-overlapping (shared with the gazetteer matcher)",
|
|
16
|
+
"feature": "emitted per-piece row = [country_surface, country_ambiguous] (the raw bits); confidence = 1.0 where country_surface fires."
|
|
17
|
+
},
|
|
18
|
+
"entries": {
|
|
19
|
+
"afghanistan": 1,
|
|
20
|
+
"åland islands": 1,
|
|
21
|
+
"albania": 1,
|
|
22
|
+
"algeria": 1,
|
|
23
|
+
"america": 3,
|
|
24
|
+
"american samoa": 1,
|
|
25
|
+
"andorra": 1,
|
|
26
|
+
"angola": 1,
|
|
27
|
+
"anguilla": 1,
|
|
28
|
+
"antarctica": 1,
|
|
29
|
+
"antigua and barbuda": 1,
|
|
30
|
+
"argentina": 1,
|
|
31
|
+
"armenia": 1,
|
|
32
|
+
"aruba": 1,
|
|
33
|
+
"australia": 1,
|
|
34
|
+
"austria": 1,
|
|
35
|
+
"azerbaijan": 1,
|
|
36
|
+
"bahamas": 1,
|
|
37
|
+
"bahrain": 1,
|
|
38
|
+
"bangladesh": 1,
|
|
39
|
+
"barbados": 1,
|
|
40
|
+
"belarus": 1,
|
|
41
|
+
"belgië": 1,
|
|
42
|
+
"belgique": 1,
|
|
43
|
+
"belgium": 1,
|
|
44
|
+
"belize": 1,
|
|
45
|
+
"benin": 1,
|
|
46
|
+
"bermuda": 1,
|
|
47
|
+
"bhutan": 1,
|
|
48
|
+
"bolivia": 1,
|
|
49
|
+
"bonaire sint eustatius and saba": 1,
|
|
50
|
+
"bosnia and herzegovina": 1,
|
|
51
|
+
"botswana": 1,
|
|
52
|
+
"bouvet island": 1,
|
|
53
|
+
"brazil": 1,
|
|
54
|
+
"britain": 3,
|
|
55
|
+
"british indian ocean territory": 1,
|
|
56
|
+
"brunei darussalam": 1,
|
|
57
|
+
"bulgaria": 1,
|
|
58
|
+
"burkina faso": 1,
|
|
59
|
+
"burundi": 1,
|
|
60
|
+
"cambodia": 1,
|
|
61
|
+
"cameroon": 1,
|
|
62
|
+
"canada": 1,
|
|
63
|
+
"cape verde": 1,
|
|
64
|
+
"cayman islands": 1,
|
|
65
|
+
"central african republic": 1,
|
|
66
|
+
"chad": 3,
|
|
67
|
+
"chile": 1,
|
|
68
|
+
"china": 1,
|
|
69
|
+
"christmas island": 1,
|
|
70
|
+
"cocos keeling islands": 1,
|
|
71
|
+
"colombia": 1,
|
|
72
|
+
"comoros": 1,
|
|
73
|
+
"congo": 1,
|
|
74
|
+
"congo the democratic republic of the": 1,
|
|
75
|
+
"cook islands": 1,
|
|
76
|
+
"costa rica": 1,
|
|
77
|
+
"cote divoire": 1,
|
|
78
|
+
"croatia": 1,
|
|
79
|
+
"cuba": 1,
|
|
80
|
+
"curaçao": 1,
|
|
81
|
+
"cyprus": 1,
|
|
82
|
+
"czech republic": 1,
|
|
83
|
+
"denmark": 1,
|
|
84
|
+
"deutschland": 1,
|
|
85
|
+
"djibouti": 1,
|
|
86
|
+
"dominica": 1,
|
|
87
|
+
"dominican republic": 1,
|
|
88
|
+
"ecuador": 1,
|
|
89
|
+
"egypt": 1,
|
|
90
|
+
"éire": 1,
|
|
91
|
+
"el salvador": 1,
|
|
92
|
+
"england": 3,
|
|
93
|
+
"equatorial guinea": 1,
|
|
94
|
+
"eritrea": 1,
|
|
95
|
+
"espana": 1,
|
|
96
|
+
"españa": 1,
|
|
97
|
+
"estonia": 1,
|
|
98
|
+
"ethiopia": 1,
|
|
99
|
+
"falkland islands malvinas": 1,
|
|
100
|
+
"faroe islands": 1,
|
|
101
|
+
"federal republic of germany": 1,
|
|
102
|
+
"fiji": 1,
|
|
103
|
+
"finland": 1,
|
|
104
|
+
"france": 1,
|
|
105
|
+
"french guiana": 1,
|
|
106
|
+
"french polynesia": 1,
|
|
107
|
+
"french republic": 1,
|
|
108
|
+
"french southern territories": 1,
|
|
109
|
+
"gabon": 1,
|
|
110
|
+
"gambia": 1,
|
|
111
|
+
"georgia": 3,
|
|
112
|
+
"germany": 1,
|
|
113
|
+
"ghana": 1,
|
|
114
|
+
"gibraltar": 1,
|
|
115
|
+
"great britain": 1,
|
|
116
|
+
"greece": 1,
|
|
117
|
+
"greenland": 1,
|
|
118
|
+
"grenada": 1,
|
|
119
|
+
"guadeloupe": 1,
|
|
120
|
+
"guam": 3,
|
|
121
|
+
"guatemala": 1,
|
|
122
|
+
"guernsey": 1,
|
|
123
|
+
"guinea": 3,
|
|
124
|
+
"guinea-bissau": 1,
|
|
125
|
+
"guyana": 1,
|
|
126
|
+
"haiti": 1,
|
|
127
|
+
"heard island and mcdonald islands": 1,
|
|
128
|
+
"holland": 1,
|
|
129
|
+
"holy see vatican city state": 1,
|
|
130
|
+
"honduras": 1,
|
|
131
|
+
"hong kong": 1,
|
|
132
|
+
"hungary": 1,
|
|
133
|
+
"iceland": 1,
|
|
134
|
+
"india": 1,
|
|
135
|
+
"indonesia": 1,
|
|
136
|
+
"iran islamic republic of": 1,
|
|
137
|
+
"iraq": 1,
|
|
138
|
+
"ireland": 1,
|
|
139
|
+
"isle of man": 1,
|
|
140
|
+
"israel": 1,
|
|
141
|
+
"italia": 1,
|
|
142
|
+
"italy": 1,
|
|
143
|
+
"jamaica": 1,
|
|
144
|
+
"japan": 1,
|
|
145
|
+
"jersey": 3,
|
|
146
|
+
"jordan": 3,
|
|
147
|
+
"kazakhstan": 1,
|
|
148
|
+
"kenya": 1,
|
|
149
|
+
"kiribati": 1,
|
|
150
|
+
"korea democratic people's republic of": 1,
|
|
151
|
+
"korea republic of": 1,
|
|
152
|
+
"kuwait": 1,
|
|
153
|
+
"kyrgyzstan": 1,
|
|
154
|
+
"lao people's democratic republic": 1,
|
|
155
|
+
"latvia": 1,
|
|
156
|
+
"lebanon": 1,
|
|
157
|
+
"lesotho": 1,
|
|
158
|
+
"liberia": 1,
|
|
159
|
+
"libya": 1,
|
|
160
|
+
"liechtenstein": 1,
|
|
161
|
+
"lithuania": 1,
|
|
162
|
+
"luxembourg": 1,
|
|
163
|
+
"macao": 1,
|
|
164
|
+
"macedonia the former yugoslav republic of": 1,
|
|
165
|
+
"madagascar": 1,
|
|
166
|
+
"malawi": 1,
|
|
167
|
+
"malaysia": 1,
|
|
168
|
+
"maldives": 1,
|
|
169
|
+
"mali": 1,
|
|
170
|
+
"malta": 1,
|
|
171
|
+
"marshall islands": 1,
|
|
172
|
+
"martinique": 1,
|
|
173
|
+
"mauritania": 1,
|
|
174
|
+
"mauritius": 1,
|
|
175
|
+
"mayotte": 1,
|
|
176
|
+
"mexico": 1,
|
|
177
|
+
"méxico": 1,
|
|
178
|
+
"micronesia federated states of": 1,
|
|
179
|
+
"moldova republic of": 1,
|
|
180
|
+
"monaco": 1,
|
|
181
|
+
"mongolia": 1,
|
|
182
|
+
"montenegro": 1,
|
|
183
|
+
"montserrat": 1,
|
|
184
|
+
"morocco": 1,
|
|
185
|
+
"mozambique": 1,
|
|
186
|
+
"myanmar": 1,
|
|
187
|
+
"namibia": 1,
|
|
188
|
+
"nauru": 1,
|
|
189
|
+
"nederland": 1,
|
|
190
|
+
"nepal": 1,
|
|
191
|
+
"netherlands": 1,
|
|
192
|
+
"new caledonia": 1,
|
|
193
|
+
"new zealand": 1,
|
|
194
|
+
"nicaragua": 1,
|
|
195
|
+
"niger": 1,
|
|
196
|
+
"nigeria": 1,
|
|
197
|
+
"nippon": 1,
|
|
198
|
+
"niue": 1,
|
|
199
|
+
"norfolk island": 1,
|
|
200
|
+
"northern mariana islands": 1,
|
|
201
|
+
"norway": 1,
|
|
202
|
+
"oman": 1,
|
|
203
|
+
"osterreich": 1,
|
|
204
|
+
"österreich": 1,
|
|
205
|
+
"pakistan": 1,
|
|
206
|
+
"palau": 1,
|
|
207
|
+
"palestine state of": 1,
|
|
208
|
+
"panama": 1,
|
|
209
|
+
"papua new guinea": 1,
|
|
210
|
+
"paraguay": 1,
|
|
211
|
+
"peru": 1,
|
|
212
|
+
"philippines": 1,
|
|
213
|
+
"pitcairn": 1,
|
|
214
|
+
"poland": 1,
|
|
215
|
+
"portugal": 1,
|
|
216
|
+
"puerto rico": 1,
|
|
217
|
+
"qatar": 1,
|
|
218
|
+
"reunion": 1,
|
|
219
|
+
"romania": 1,
|
|
220
|
+
"russian federation": 1,
|
|
221
|
+
"rwanda": 1,
|
|
222
|
+
"saint barthélemy": 1,
|
|
223
|
+
"saint helena ascension and tristan da cunha": 1,
|
|
224
|
+
"saint kitts and nevis": 1,
|
|
225
|
+
"saint lucia": 1,
|
|
226
|
+
"saint martin french part": 1,
|
|
227
|
+
"saint pierre and miquelon": 1,
|
|
228
|
+
"saint vincent and the grenadines": 1,
|
|
229
|
+
"samoa": 1,
|
|
230
|
+
"san marino": 1,
|
|
231
|
+
"sao tome and principe": 1,
|
|
232
|
+
"saudi arabia": 1,
|
|
233
|
+
"schweiz": 1,
|
|
234
|
+
"senegal": 1,
|
|
235
|
+
"serbia": 1,
|
|
236
|
+
"seychelles": 1,
|
|
237
|
+
"sierra leone": 1,
|
|
238
|
+
"singapore": 1,
|
|
239
|
+
"sint maarten dutch part": 1,
|
|
240
|
+
"slovakia": 1,
|
|
241
|
+
"slovenia": 1,
|
|
242
|
+
"solomon islands": 1,
|
|
243
|
+
"somalia": 1,
|
|
244
|
+
"south africa": 1,
|
|
245
|
+
"south georgia and the south sandwich islands": 1,
|
|
246
|
+
"south sudan": 1,
|
|
247
|
+
"spain": 1,
|
|
248
|
+
"sri lanka": 1,
|
|
249
|
+
"sudan": 1,
|
|
250
|
+
"suisse": 1,
|
|
251
|
+
"suriname": 1,
|
|
252
|
+
"svalbard and jan mayen": 1,
|
|
253
|
+
"svizzera": 1,
|
|
254
|
+
"swaziland": 1,
|
|
255
|
+
"sweden": 1,
|
|
256
|
+
"switzerland": 1,
|
|
257
|
+
"syrian arab republic": 1,
|
|
258
|
+
"taiwan province of china": 1,
|
|
259
|
+
"tajikistan": 1,
|
|
260
|
+
"tanzania united republic of": 1,
|
|
261
|
+
"thailand": 1,
|
|
262
|
+
"the netherlands": 1,
|
|
263
|
+
"timor-leste": 1,
|
|
264
|
+
"togo": 1,
|
|
265
|
+
"tokelau": 1,
|
|
266
|
+
"tonga": 1,
|
|
267
|
+
"trinidad and tobago": 1,
|
|
268
|
+
"tunisia": 1,
|
|
269
|
+
"turkey": 3,
|
|
270
|
+
"turkmenistan": 1,
|
|
271
|
+
"turks and caicos islands": 1,
|
|
272
|
+
"tuvalu": 1,
|
|
273
|
+
"uganda": 1,
|
|
274
|
+
"ukraine": 1,
|
|
275
|
+
"united arab emirates": 1,
|
|
276
|
+
"united kingdom": 1,
|
|
277
|
+
"united states": 1,
|
|
278
|
+
"united states minor outlying islands": 1,
|
|
279
|
+
"united states of america": 1,
|
|
280
|
+
"uruguay": 1,
|
|
281
|
+
"uzbekistan": 1,
|
|
282
|
+
"vanuatu": 1,
|
|
283
|
+
"venezuela": 1,
|
|
284
|
+
"viet nam": 1,
|
|
285
|
+
"virgin islands british": 1,
|
|
286
|
+
"virgin islands u.s": 1,
|
|
287
|
+
"wallis and futuna": 1,
|
|
288
|
+
"western sahara": 1,
|
|
289
|
+
"yemen": 1,
|
|
290
|
+
"zambia": 1,
|
|
291
|
+
"zimbabwe": 1
|
|
292
|
+
},
|
|
293
|
+
"code_entries": {
|
|
294
|
+
"AT": 1,
|
|
295
|
+
"AU": 1,
|
|
296
|
+
"AUS": 1,
|
|
297
|
+
"BE": 1,
|
|
298
|
+
"CA": 3,
|
|
299
|
+
"CAN": 1,
|
|
300
|
+
"CH": 1,
|
|
301
|
+
"DE": 3,
|
|
302
|
+
"ES": 1,
|
|
303
|
+
"ESP": 1,
|
|
304
|
+
"FR": 1,
|
|
305
|
+
"FRA": 1,
|
|
306
|
+
"GB": 1,
|
|
307
|
+
"GER": 1,
|
|
308
|
+
"IE": 1,
|
|
309
|
+
"IRL": 1,
|
|
310
|
+
"IT": 1,
|
|
311
|
+
"ITA": 1,
|
|
312
|
+
"JP": 1,
|
|
313
|
+
"JPN": 1,
|
|
314
|
+
"MEX": 1,
|
|
315
|
+
"MX": 1,
|
|
316
|
+
"NL": 1,
|
|
317
|
+
"NLD": 1,
|
|
318
|
+
"U.K": 1,
|
|
319
|
+
"U.S": 1,
|
|
320
|
+
"U.S.A": 1,
|
|
321
|
+
"UK": 1,
|
|
322
|
+
"US": 1,
|
|
323
|
+
"USA": 1,
|
|
324
|
+
"日本": 1
|
|
325
|
+
}
|
|
326
|
+
}
|
package/model-card.json
CHANGED
|
@@ -1,25 +1,28 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "neural-weights-fr-fr",
|
|
3
|
-
"version": "
|
|
4
|
-
"model_lineage": "Ships the en-us
|
|
5
|
-
"phase": "Stage 3 — multi-locale (FR served by the shared en-us
|
|
3
|
+
"version": "6.2.0",
|
|
4
|
+
"model_lineage": "Ships the en-us v263 (v2.6.3-country-channel, #1104) model — init_from v261 + the dedicated country soft-feed channel; tokenizer v0.9.0-multisplice (73,143). One model serves both locales; see the en-us card for the full lineage, the v263 grade, and the documented coordinate-invisible region trade.",
|
|
5
|
+
"phase": "Stage 3 — multi-locale (FR served by the shared en-us v263 model, shipped as 6.2.0)",
|
|
6
6
|
"license": "AGPL-3.0-only OR LicenseRef-Commercial",
|
|
7
7
|
"locale": "fr-fr",
|
|
8
8
|
"training": {
|
|
9
|
-
"corpus_version": "
|
|
10
|
-
"tokenizer_version": "0.
|
|
11
|
-
"steps":
|
|
9
|
+
"corpus_version": "v257 fragment-campaign corpus (shard-v5) re-tokenized through v0.9.0-multisplice; span-boundary aux head from BIO labels.",
|
|
10
|
+
"tokenizer_version": "0.9.0-multisplice",
|
|
11
|
+
"steps": 8000,
|
|
12
12
|
"hardware": "NVIDIA A100-SXM4-40GB (Modal cloud)",
|
|
13
|
-
"note": "Identical artifact to neural-weights-en-us
|
|
13
|
+
"note": "Identical artifact to neural-weights-en-us 6.2.0 / v263 country-channel (one multi-locale model serves both, copied at publish time). See that card's training/recipe for full provenance."
|
|
14
14
|
},
|
|
15
15
|
"requires": {
|
|
16
|
-
"$comment": "SHIP-CONFIG mirrored from neural-weights-en-us (#718/#721): this package ships en-us's model, so its required channels are identical. createScorer / loadFromWeights read this and feed anchor + gazetteer + the conventions mask; running channel-OFF is the #566/#685 OOD crater.
|
|
16
|
+
"$comment": "SHIP-CONFIG mirrored from neural-weights-en-us (#718/#721): this package ships en-us's model, so its required channels are identical — including `country` (the #1104 dedicated country soft-feed channel added in v263). createScorer / loadFromWeights read this and feed anchor + gazetteer + country + the conventions mask; running channel-OFF is the #566/#685 OOD crater.",
|
|
17
17
|
"anchor": {
|
|
18
18
|
"required": true
|
|
19
19
|
},
|
|
20
20
|
"gazetteer": {
|
|
21
21
|
"required": true
|
|
22
22
|
},
|
|
23
|
+
"country": {
|
|
24
|
+
"required": true
|
|
25
|
+
},
|
|
23
26
|
"conventions": {
|
|
24
27
|
"required": true,
|
|
25
28
|
"mode": "auto"
|
package/model.onnx
CHANGED
|
Binary file
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@mailwoman/neural-weights-fr-fr",
|
|
3
|
-
"version": "6.
|
|
3
|
+
"version": "6.2.0",
|
|
4
4
|
"description": "Mailwoman neural-classifier weights for locale 'fr-fr'. Data-only package — loaded by @mailwoman/neural at runtime.",
|
|
5
5
|
"license": "AGPL-3.0-only OR LicenseRef-Commercial",
|
|
6
6
|
"repository": {
|
|
@@ -14,6 +14,7 @@
|
|
|
14
14
|
"model-card.json",
|
|
15
15
|
"postcode-fr.bin",
|
|
16
16
|
"anchor-lexicon-v1.json",
|
|
17
|
+
"country-surface-lexicon-v1.json",
|
|
17
18
|
"README.md"
|
|
18
19
|
],
|
|
19
20
|
"publishConfig": {
|
package/tokenizer.model
CHANGED
|
Binary file
|