Demosthene-OR
commited on
Commit
•
e753237
1
Parent(s):
4764028
Ajout des fichiers vocab_*.txt
Browse files- data/vocab_en.txt +201 -0
- data/vocab_fr.txt +340 -0
- main_dl.py +0 -1
data/vocab_en.txt
ADDED
@@ -0,0 +1,201 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
is
|
2 |
+
in
|
3 |
+
it
|
4 |
+
during
|
5 |
+
the
|
6 |
+
but
|
7 |
+
and
|
8 |
+
sometimes
|
9 |
+
usually
|
10 |
+
never
|
11 |
+
favorite
|
12 |
+
least
|
13 |
+
fruit
|
14 |
+
most
|
15 |
+
loved
|
16 |
+
liked
|
17 |
+
new
|
18 |
+
paris
|
19 |
+
jersey
|
20 |
+
california
|
21 |
+
india
|
22 |
+
united
|
23 |
+
states
|
24 |
+
france
|
25 |
+
china
|
26 |
+
she
|
27 |
+
he
|
28 |
+
grapefruit
|
29 |
+
her
|
30 |
+
his
|
31 |
+
my
|
32 |
+
your
|
33 |
+
january
|
34 |
+
winter
|
35 |
+
summer
|
36 |
+
october
|
37 |
+
march
|
38 |
+
autumn
|
39 |
+
september
|
40 |
+
july
|
41 |
+
june
|
42 |
+
may
|
43 |
+
fall
|
44 |
+
spring
|
45 |
+
nice
|
46 |
+
their
|
47 |
+
snowy
|
48 |
+
freezing
|
49 |
+
december
|
50 |
+
our
|
51 |
+
april
|
52 |
+
november
|
53 |
+
pleasant
|
54 |
+
warm
|
55 |
+
beautiful
|
56 |
+
february
|
57 |
+
busy
|
58 |
+
dry
|
59 |
+
cold
|
60 |
+
chilly
|
61 |
+
rainy
|
62 |
+
august
|
63 |
+
relaxing
|
64 |
+
wonderful
|
65 |
+
mild
|
66 |
+
quiet
|
67 |
+
wet
|
68 |
+
hot
|
69 |
+
likes
|
70 |
+
dislikes
|
71 |
+
peaches
|
72 |
+
lemons
|
73 |
+
grapes
|
74 |
+
bananas
|
75 |
+
oranges
|
76 |
+
pears
|
77 |
+
limes
|
78 |
+
strawberries
|
79 |
+
apples
|
80 |
+
mangoes
|
81 |
+
to
|
82 |
+
banana
|
83 |
+
mango
|
84 |
+
pear
|
85 |
+
apple
|
86 |
+
orange
|
87 |
+
lime
|
88 |
+
grape
|
89 |
+
peach
|
90 |
+
lemon
|
91 |
+
strawberry
|
92 |
+
like
|
93 |
+
dislike
|
94 |
+
they
|
95 |
+
that
|
96 |
+
i
|
97 |
+
we
|
98 |
+
you
|
99 |
+
animal
|
100 |
+
truck
|
101 |
+
a
|
102 |
+
automobile
|
103 |
+
car
|
104 |
+
was
|
105 |
+
next
|
106 |
+
go
|
107 |
+
driving
|
108 |
+
visit
|
109 |
+
big
|
110 |
+
little
|
111 |
+
yellow
|
112 |
+
shiny
|
113 |
+
green
|
114 |
+
black
|
115 |
+
blue
|
116 |
+
white
|
117 |
+
rusty
|
118 |
+
red
|
119 |
+
old
|
120 |
+
are
|
121 |
+
last
|
122 |
+
this
|
123 |
+
animals
|
124 |
+
feared
|
125 |
+
plan
|
126 |
+
going
|
127 |
+
drove
|
128 |
+
disliked
|
129 |
+
drives
|
130 |
+
saw
|
131 |
+
between
|
132 |
+
translate
|
133 |
+
plans
|
134 |
+
went
|
135 |
+
were
|
136 |
+
wanted
|
137 |
+
thinks
|
138 |
+
might
|
139 |
+
french
|
140 |
+
spanish
|
141 |
+
chinese
|
142 |
+
portuguese
|
143 |
+
english
|
144 |
+
translating
|
145 |
+
difficult
|
146 |
+
wants
|
147 |
+
fun
|
148 |
+
easy
|
149 |
+
its
|
150 |
+
why
|
151 |
+
think
|
152 |
+
bear
|
153 |
+
did
|
154 |
+
mouse
|
155 |
+
bird
|
156 |
+
cat
|
157 |
+
rabbit
|
158 |
+
shark
|
159 |
+
lion
|
160 |
+
snake
|
161 |
+
horse
|
162 |
+
monkey
|
163 |
+
elephant
|
164 |
+
dog
|
165 |
+
when
|
166 |
+
want
|
167 |
+
do
|
168 |
+
horses
|
169 |
+
snakes
|
170 |
+
sharks
|
171 |
+
didnt
|
172 |
+
birds
|
173 |
+
monkeys
|
174 |
+
elephants
|
175 |
+
cats
|
176 |
+
how
|
177 |
+
rabbits
|
178 |
+
bears
|
179 |
+
mice
|
180 |
+
lions
|
181 |
+
tower
|
182 |
+
store
|
183 |
+
grocery
|
184 |
+
eiffel
|
185 |
+
dogs
|
186 |
+
school
|
187 |
+
lake
|
188 |
+
football
|
189 |
+
field
|
190 |
+
would
|
191 |
+
been
|
192 |
+
arent
|
193 |
+
weather
|
194 |
+
has
|
195 |
+
does
|
196 |
+
isnt
|
197 |
+
am
|
198 |
+
have
|
199 |
+
where
|
200 |
+
[start]
|
201 |
+
[end]
|
data/vocab_fr.txt
ADDED
@@ -0,0 +1,340 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
est
|
2 |
+
en
|
3 |
+
il
|
4 |
+
les
|
5 |
+
mais
|
6 |
+
et
|
7 |
+
la
|
8 |
+
parfois
|
9 |
+
jamais
|
10 |
+
le
|
11 |
+
l
|
12 |
+
généralement
|
13 |
+
moins
|
14 |
+
aimé
|
15 |
+
au
|
16 |
+
fruit
|
17 |
+
préféré
|
18 |
+
agréable
|
19 |
+
froid
|
20 |
+
son
|
21 |
+
chaud
|
22 |
+
a
|
23 |
+
de
|
24 |
+
plus
|
25 |
+
automne
|
26 |
+
mois
|
27 |
+
elle
|
28 |
+
citrons
|
29 |
+
paris
|
30 |
+
inde
|
31 |
+
étatsunis
|
32 |
+
jersey
|
33 |
+
france
|
34 |
+
new
|
35 |
+
chine
|
36 |
+
pendant
|
37 |
+
pamplemousse
|
38 |
+
mon
|
39 |
+
votre
|
40 |
+
janvier
|
41 |
+
été
|
42 |
+
hiver
|
43 |
+
octobre
|
44 |
+
mars
|
45 |
+
septembre
|
46 |
+
juillet
|
47 |
+
juin
|
48 |
+
mai
|
49 |
+
printemps
|
50 |
+
décembre
|
51 |
+
avril
|
52 |
+
novembre
|
53 |
+
février
|
54 |
+
aime
|
55 |
+
août
|
56 |
+
merveilleux
|
57 |
+
relaxant
|
58 |
+
doux
|
59 |
+
humide
|
60 |
+
notre
|
61 |
+
californie
|
62 |
+
sec
|
63 |
+
leur
|
64 |
+
occupé
|
65 |
+
pluvieux
|
66 |
+
calme
|
67 |
+
beau
|
68 |
+
habituellement
|
69 |
+
pêches
|
70 |
+
bananes
|
71 |
+
oranges
|
72 |
+
poires
|
73 |
+
raisins
|
74 |
+
verts
|
75 |
+
fraises
|
76 |
+
pommes
|
77 |
+
mangues
|
78 |
+
d
|
79 |
+
mangue
|
80 |
+
gel
|
81 |
+
banane
|
82 |
+
poire
|
83 |
+
pomme
|
84 |
+
lorange
|
85 |
+
chaux
|
86 |
+
raisin
|
87 |
+
pêche
|
88 |
+
citron
|
89 |
+
fraise
|
90 |
+
pas
|
91 |
+
enneigée
|
92 |
+
favori
|
93 |
+
déteste
|
94 |
+
fruits
|
95 |
+
gèle
|
96 |
+
voiture
|
97 |
+
lautomne
|
98 |
+
ils
|
99 |
+
naime
|
100 |
+
california
|
101 |
+
neige
|
102 |
+
fait
|
103 |
+
ne
|
104 |
+
belle
|
105 |
+
vous
|
106 |
+
nous
|
107 |
+
des
|
108 |
+
animal
|
109 |
+
camion
|
110 |
+
cours
|
111 |
+
neigeux
|
112 |
+
conduit
|
113 |
+
prochain
|
114 |
+
ce
|
115 |
+
je
|
116 |
+
tranquille
|
117 |
+
cher
|
118 |
+
une
|
119 |
+
cette
|
120 |
+
aller
|
121 |
+
était
|
122 |
+
naiment
|
123 |
+
chaude
|
124 |
+
aimons
|
125 |
+
aiment
|
126 |
+
naimez
|
127 |
+
aimez
|
128 |
+
leurs
|
129 |
+
sont
|
130 |
+
détestons
|
131 |
+
jaune
|
132 |
+
jaime
|
133 |
+
visiter
|
134 |
+
rouge
|
135 |
+
occupée
|
136 |
+
sèche
|
137 |
+
frisquet
|
138 |
+
dernier
|
139 |
+
animaux
|
140 |
+
préférée
|
141 |
+
un
|
142 |
+
aimait
|
143 |
+
que
|
144 |
+
conduisait
|
145 |
+
vu
|
146 |
+
verte
|
147 |
+
nos
|
148 |
+
nouvelle
|
149 |
+
vieille
|
150 |
+
brillant
|
151 |
+
petite
|
152 |
+
noire
|
153 |
+
redouté
|
154 |
+
blanche
|
155 |
+
naimait
|
156 |
+
pense
|
157 |
+
pleut
|
158 |
+
pamplemousses
|
159 |
+
entre
|
160 |
+
nouveau
|
161 |
+
traduire
|
162 |
+
bleu
|
163 |
+
grande
|
164 |
+
bleue
|
165 |
+
rouillée
|
166 |
+
rouillé
|
167 |
+
se
|
168 |
+
ses
|
169 |
+
quil
|
170 |
+
aux
|
171 |
+
blanc
|
172 |
+
préférés
|
173 |
+
pluies
|
174 |
+
brillante
|
175 |
+
va
|
176 |
+
noir
|
177 |
+
étaient
|
178 |
+
envisage
|
179 |
+
vert
|
180 |
+
rendre
|
181 |
+
petit
|
182 |
+
français
|
183 |
+
espagnol
|
184 |
+
glaciales
|
185 |
+
vieux
|
186 |
+
chinois
|
187 |
+
portugais
|
188 |
+
mes
|
189 |
+
anglais
|
190 |
+
cet
|
191 |
+
comme
|
192 |
+
automobile
|
193 |
+
mouillé
|
194 |
+
gros
|
195 |
+
difficile
|
196 |
+
traduction
|
197 |
+
souris
|
198 |
+
veut
|
199 |
+
amusant
|
200 |
+
voulait
|
201 |
+
facile
|
202 |
+
pourquoi
|
203 |
+
prévoyons
|
204 |
+
aimés
|
205 |
+
prévois
|
206 |
+
pourrait
|
207 |
+
intention
|
208 |
+
vos
|
209 |
+
chat
|
210 |
+
ours
|
211 |
+
allé
|
212 |
+
ont
|
213 |
+
lapin
|
214 |
+
clémentes
|
215 |
+
requin
|
216 |
+
lion
|
217 |
+
serpent
|
218 |
+
redoutés
|
219 |
+
cheval
|
220 |
+
singe
|
221 |
+
maillot
|
222 |
+
avez
|
223 |
+
trop
|
224 |
+
monde
|
225 |
+
grosse
|
226 |
+
i
|
227 |
+
chien
|
228 |
+
pluie
|
229 |
+
vont
|
230 |
+
volant
|
231 |
+
allés
|
232 |
+
allée
|
233 |
+
quand
|
234 |
+
oiseau
|
235 |
+
veulent
|
236 |
+
voulaient
|
237 |
+
éléphant
|
238 |
+
pourraient
|
239 |
+
magnifique
|
240 |
+
aimée
|
241 |
+
gelé
|
242 |
+
lautomobile
|
243 |
+
détendre
|
244 |
+
naimons
|
245 |
+
grand
|
246 |
+
détestait
|
247 |
+
prévoit
|
248 |
+
lui
|
249 |
+
bien
|
250 |
+
vers
|
251 |
+
chevaux
|
252 |
+
serpents
|
253 |
+
visite
|
254 |
+
prévoient
|
255 |
+
lours
|
256 |
+
requins
|
257 |
+
oiseaux
|
258 |
+
loiseau
|
259 |
+
lanimal
|
260 |
+
éléphants
|
261 |
+
singes
|
262 |
+
légère
|
263 |
+
chats
|
264 |
+
cépage
|
265 |
+
comment
|
266 |
+
lapins
|
267 |
+
léléphant
|
268 |
+
lions
|
269 |
+
tour
|
270 |
+
lépicerie
|
271 |
+
eiffel
|
272 |
+
chiens
|
273 |
+
lécole
|
274 |
+
lac
|
275 |
+
terrain
|
276 |
+
football
|
277 |
+
pensez
|
278 |
+
allons
|
279 |
+
pousse
|
280 |
+
nest
|
281 |
+
du
|
282 |
+
allez
|
283 |
+
peu
|
284 |
+
rouille
|
285 |
+
êtesvous
|
286 |
+
temps
|
287 |
+
at
|
288 |
+
quelle
|
289 |
+
sur
|
290 |
+
petites
|
291 |
+
voudrait
|
292 |
+
frais
|
293 |
+
dernière
|
294 |
+
cest
|
295 |
+
avons
|
296 |
+
t
|
297 |
+
proches
|
298 |
+
vais
|
299 |
+
manguiers
|
300 |
+
détestez
|
301 |
+
préférées
|
302 |
+
dans
|
303 |
+
porcelaine
|
304 |
+
grandes
|
305 |
+
voulez
|
306 |
+
plaît
|
307 |
+
petits
|
308 |
+
na
|
309 |
+
congélation
|
310 |
+
estce
|
311 |
+
envisagent
|
312 |
+
aimeraient
|
313 |
+
où
|
314 |
+
limes
|
315 |
+
durant
|
316 |
+
aimentils
|
317 |
+
douce
|
318 |
+
conduite
|
319 |
+
grands
|
320 |
+
gelés
|
321 |
+
enneigé
|
322 |
+
moindres
|
323 |
+
grosses
|
324 |
+
tout
|
325 |
+
bénigne
|
326 |
+
traduis
|
327 |
+
nêtes
|
328 |
+
etatsunis
|
329 |
+
vit
|
330 |
+
trouvé
|
331 |
+
souvent
|
332 |
+
ressort
|
333 |
+
redoutée
|
334 |
+
qui
|
335 |
+
mouillée
|
336 |
+
moteur
|
337 |
+
favoris
|
338 |
+
détend
|
339 |
+
[start]
|
340 |
+
[end]
|
main_dl.py
CHANGED
@@ -30,7 +30,6 @@ def custom_standardization(input_string):
|
|
30 |
return tf.strings.regex_replace(
|
31 |
lowercase, f"[{re.escape(strip_chars)}]", "")
|
32 |
|
33 |
-
@st.cache_data
|
34 |
def load_vocab(file_path):
|
35 |
with open(file_path, "r", encoding="utf-8") as file:
|
36 |
return file.read().split('\n')[:-1]
|
|
|
30 |
return tf.strings.regex_replace(
|
31 |
lowercase, f"[{re.escape(strip_chars)}]", "")
|
32 |
|
|
|
33 |
def load_vocab(file_path):
|
34 |
with open(file_path, "r", encoding="utf-8") as file:
|
35 |
return file.read().split('\n')[:-1]
|