End of training
Browse files- added_tokens.json +16 -0
- special_tokens_map.json +16 -2
- tokenizer.json +148 -4
- tokenizer_config.json +148 -6
added_tokens.json
CHANGED
@@ -1,10 +1,26 @@
|
|
1 |
{
|
2 |
"</s>": 2,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
"<mask>": 57521,
|
4 |
"<pad>": 1,
|
5 |
"<s>": 0,
|
|
|
|
|
|
|
|
|
|
|
6 |
"<s_iitcdip>": 57523,
|
|
|
7 |
"<s_synthdog>": 57524,
|
|
|
|
|
8 |
"<sep/>": 57522,
|
9 |
"<unk>": 3
|
10 |
}
|
|
|
1 |
{
|
2 |
"</s>": 2,
|
3 |
+
"</s_address>": 57528,
|
4 |
+
"</s_bill>": 57526,
|
5 |
+
"</s_company>": 57533,
|
6 |
+
"</s_date>": 57527,
|
7 |
+
"</s_description>": 57540,
|
8 |
+
"</s_item no>": 57537,
|
9 |
+
"</s_total amount>": 57529,
|
10 |
+
"</s_total>": 57530,
|
11 |
"<mask>": 57521,
|
12 |
"<pad>": 1,
|
13 |
"<s>": 0,
|
14 |
+
"<s_address>": 57536,
|
15 |
+
"<s_bill>": 57534,
|
16 |
+
"<s_company>": 57539,
|
17 |
+
"<s_date>": 57525,
|
18 |
+
"<s_description>": 57535,
|
19 |
"<s_iitcdip>": 57523,
|
20 |
+
"<s_item no>": 57532,
|
21 |
"<s_synthdog>": 57524,
|
22 |
+
"<s_total amount>": 57531,
|
23 |
+
"<s_total>": 57538,
|
24 |
"<sep/>": 57522,
|
25 |
"<unk>": 3
|
26 |
}
|
special_tokens_map.json
CHANGED
@@ -2,8 +2,22 @@
|
|
2 |
"additional_special_tokens": [
|
3 |
"<s_iitcdip>",
|
4 |
"<s_synthdog>",
|
5 |
-
"
|
6 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
],
|
8 |
"bos_token": "<s>",
|
9 |
"cls_token": "<s>",
|
|
|
2 |
"additional_special_tokens": [
|
3 |
"<s_iitcdip>",
|
4 |
"<s_synthdog>",
|
5 |
+
"<s_date>",
|
6 |
+
"</s_bill>",
|
7 |
+
"</s_date>",
|
8 |
+
"</s_address>",
|
9 |
+
"</s_total amount>",
|
10 |
+
"</s_total>",
|
11 |
+
"<s_total amount>",
|
12 |
+
"<s_item no>",
|
13 |
+
"</s_company>",
|
14 |
+
"<s_bill>",
|
15 |
+
"<s_description>",
|
16 |
+
"<s_address>",
|
17 |
+
"</s_item no>",
|
18 |
+
"<s_total>",
|
19 |
+
"<s_company>",
|
20 |
+
"</s_description>"
|
21 |
],
|
22 |
"bos_token": "<s>",
|
23 |
"cls_token": "<s>",
|
tokenizer.json
CHANGED
@@ -21,8 +21,8 @@
|
|
21 |
"id": 0,
|
22 |
"content": "<s>",
|
23 |
"single_word": false,
|
24 |
-
"lstrip":
|
25 |
-
"rstrip":
|
26 |
"normalized": false,
|
27 |
"special": true
|
28 |
},
|
@@ -39,8 +39,8 @@
|
|
39 |
"id": 2,
|
40 |
"content": "</s>",
|
41 |
"single_word": false,
|
42 |
-
"lstrip":
|
43 |
-
"rstrip":
|
44 |
"normalized": false,
|
45 |
"special": true
|
46 |
},
|
@@ -88,6 +88,150 @@
|
|
88 |
"rstrip": true,
|
89 |
"normalized": false,
|
90 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
91 |
}
|
92 |
],
|
93 |
"normalizer": {
|
|
|
21 |
"id": 0,
|
22 |
"content": "<s>",
|
23 |
"single_word": false,
|
24 |
+
"lstrip": false,
|
25 |
+
"rstrip": false,
|
26 |
"normalized": false,
|
27 |
"special": true
|
28 |
},
|
|
|
39 |
"id": 2,
|
40 |
"content": "</s>",
|
41 |
"single_word": false,
|
42 |
+
"lstrip": false,
|
43 |
+
"rstrip": false,
|
44 |
"normalized": false,
|
45 |
"special": true
|
46 |
},
|
|
|
88 |
"rstrip": true,
|
89 |
"normalized": false,
|
90 |
"special": true
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"id": 57525,
|
94 |
+
"content": "<s_date>",
|
95 |
+
"single_word": false,
|
96 |
+
"lstrip": true,
|
97 |
+
"rstrip": true,
|
98 |
+
"normalized": false,
|
99 |
+
"special": true
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"id": 57526,
|
103 |
+
"content": "</s_bill>",
|
104 |
+
"single_word": false,
|
105 |
+
"lstrip": true,
|
106 |
+
"rstrip": true,
|
107 |
+
"normalized": false,
|
108 |
+
"special": true
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"id": 57527,
|
112 |
+
"content": "</s_date>",
|
113 |
+
"single_word": false,
|
114 |
+
"lstrip": true,
|
115 |
+
"rstrip": true,
|
116 |
+
"normalized": false,
|
117 |
+
"special": true
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"id": 57528,
|
121 |
+
"content": "</s_address>",
|
122 |
+
"single_word": false,
|
123 |
+
"lstrip": true,
|
124 |
+
"rstrip": true,
|
125 |
+
"normalized": false,
|
126 |
+
"special": true
|
127 |
+
},
|
128 |
+
{
|
129 |
+
"id": 57529,
|
130 |
+
"content": "</s_total amount>",
|
131 |
+
"single_word": false,
|
132 |
+
"lstrip": true,
|
133 |
+
"rstrip": true,
|
134 |
+
"normalized": false,
|
135 |
+
"special": true
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"id": 57530,
|
139 |
+
"content": "</s_total>",
|
140 |
+
"single_word": false,
|
141 |
+
"lstrip": true,
|
142 |
+
"rstrip": true,
|
143 |
+
"normalized": false,
|
144 |
+
"special": true
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"id": 57531,
|
148 |
+
"content": "<s_total amount>",
|
149 |
+
"single_word": false,
|
150 |
+
"lstrip": true,
|
151 |
+
"rstrip": true,
|
152 |
+
"normalized": false,
|
153 |
+
"special": true
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"id": 57532,
|
157 |
+
"content": "<s_item no>",
|
158 |
+
"single_word": false,
|
159 |
+
"lstrip": true,
|
160 |
+
"rstrip": true,
|
161 |
+
"normalized": false,
|
162 |
+
"special": true
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"id": 57533,
|
166 |
+
"content": "</s_company>",
|
167 |
+
"single_word": false,
|
168 |
+
"lstrip": true,
|
169 |
+
"rstrip": true,
|
170 |
+
"normalized": false,
|
171 |
+
"special": true
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"id": 57534,
|
175 |
+
"content": "<s_bill>",
|
176 |
+
"single_word": false,
|
177 |
+
"lstrip": true,
|
178 |
+
"rstrip": true,
|
179 |
+
"normalized": false,
|
180 |
+
"special": true
|
181 |
+
},
|
182 |
+
{
|
183 |
+
"id": 57535,
|
184 |
+
"content": "<s_description>",
|
185 |
+
"single_word": false,
|
186 |
+
"lstrip": true,
|
187 |
+
"rstrip": true,
|
188 |
+
"normalized": false,
|
189 |
+
"special": true
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"id": 57536,
|
193 |
+
"content": "<s_address>",
|
194 |
+
"single_word": false,
|
195 |
+
"lstrip": true,
|
196 |
+
"rstrip": true,
|
197 |
+
"normalized": false,
|
198 |
+
"special": true
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"id": 57537,
|
202 |
+
"content": "</s_item no>",
|
203 |
+
"single_word": false,
|
204 |
+
"lstrip": true,
|
205 |
+
"rstrip": true,
|
206 |
+
"normalized": false,
|
207 |
+
"special": true
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"id": 57538,
|
211 |
+
"content": "<s_total>",
|
212 |
+
"single_word": false,
|
213 |
+
"lstrip": true,
|
214 |
+
"rstrip": true,
|
215 |
+
"normalized": false,
|
216 |
+
"special": true
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"id": 57539,
|
220 |
+
"content": "<s_company>",
|
221 |
+
"single_word": false,
|
222 |
+
"lstrip": true,
|
223 |
+
"rstrip": true,
|
224 |
+
"normalized": false,
|
225 |
+
"special": true
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"id": 57540,
|
229 |
+
"content": "</s_description>",
|
230 |
+
"single_word": false,
|
231 |
+
"lstrip": true,
|
232 |
+
"rstrip": true,
|
233 |
+
"normalized": false,
|
234 |
+
"special": true
|
235 |
}
|
236 |
],
|
237 |
"normalizer": {
|
tokenizer_config.json
CHANGED
@@ -2,9 +2,9 @@
|
|
2 |
"added_tokens_decoder": {
|
3 |
"0": {
|
4 |
"content": "<s>",
|
5 |
-
"lstrip":
|
6 |
"normalized": false,
|
7 |
-
"rstrip":
|
8 |
"single_word": false,
|
9 |
"special": true
|
10 |
},
|
@@ -18,9 +18,9 @@
|
|
18 |
},
|
19 |
"2": {
|
20 |
"content": "</s>",
|
21 |
-
"lstrip":
|
22 |
"normalized": false,
|
23 |
-
"rstrip":
|
24 |
"single_word": false,
|
25 |
"special": true
|
26 |
},
|
@@ -63,13 +63,155 @@
|
|
63 |
"rstrip": true,
|
64 |
"single_word": false,
|
65 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
66 |
}
|
67 |
},
|
68 |
"additional_special_tokens": [
|
69 |
"<s_iitcdip>",
|
70 |
"<s_synthdog>",
|
71 |
-
"
|
72 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
73 |
],
|
74 |
"bos_token": "<s>",
|
75 |
"clean_up_tokenization_spaces": true,
|
|
|
2 |
"added_tokens_decoder": {
|
3 |
"0": {
|
4 |
"content": "<s>",
|
5 |
+
"lstrip": false,
|
6 |
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
"single_word": false,
|
9 |
"special": true
|
10 |
},
|
|
|
18 |
},
|
19 |
"2": {
|
20 |
"content": "</s>",
|
21 |
+
"lstrip": false,
|
22 |
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
"single_word": false,
|
25 |
"special": true
|
26 |
},
|
|
|
63 |
"rstrip": true,
|
64 |
"single_word": false,
|
65 |
"special": true
|
66 |
+
},
|
67 |
+
"57525": {
|
68 |
+
"content": "<s_date>",
|
69 |
+
"lstrip": true,
|
70 |
+
"normalized": false,
|
71 |
+
"rstrip": true,
|
72 |
+
"single_word": false,
|
73 |
+
"special": true
|
74 |
+
},
|
75 |
+
"57526": {
|
76 |
+
"content": "</s_bill>",
|
77 |
+
"lstrip": true,
|
78 |
+
"normalized": false,
|
79 |
+
"rstrip": true,
|
80 |
+
"single_word": false,
|
81 |
+
"special": true
|
82 |
+
},
|
83 |
+
"57527": {
|
84 |
+
"content": "</s_date>",
|
85 |
+
"lstrip": true,
|
86 |
+
"normalized": false,
|
87 |
+
"rstrip": true,
|
88 |
+
"single_word": false,
|
89 |
+
"special": true
|
90 |
+
},
|
91 |
+
"57528": {
|
92 |
+
"content": "</s_address>",
|
93 |
+
"lstrip": true,
|
94 |
+
"normalized": false,
|
95 |
+
"rstrip": true,
|
96 |
+
"single_word": false,
|
97 |
+
"special": true
|
98 |
+
},
|
99 |
+
"57529": {
|
100 |
+
"content": "</s_total amount>",
|
101 |
+
"lstrip": true,
|
102 |
+
"normalized": false,
|
103 |
+
"rstrip": true,
|
104 |
+
"single_word": false,
|
105 |
+
"special": true
|
106 |
+
},
|
107 |
+
"57530": {
|
108 |
+
"content": "</s_total>",
|
109 |
+
"lstrip": true,
|
110 |
+
"normalized": false,
|
111 |
+
"rstrip": true,
|
112 |
+
"single_word": false,
|
113 |
+
"special": true
|
114 |
+
},
|
115 |
+
"57531": {
|
116 |
+
"content": "<s_total amount>",
|
117 |
+
"lstrip": true,
|
118 |
+
"normalized": false,
|
119 |
+
"rstrip": true,
|
120 |
+
"single_word": false,
|
121 |
+
"special": true
|
122 |
+
},
|
123 |
+
"57532": {
|
124 |
+
"content": "<s_item no>",
|
125 |
+
"lstrip": true,
|
126 |
+
"normalized": false,
|
127 |
+
"rstrip": true,
|
128 |
+
"single_word": false,
|
129 |
+
"special": true
|
130 |
+
},
|
131 |
+
"57533": {
|
132 |
+
"content": "</s_company>",
|
133 |
+
"lstrip": true,
|
134 |
+
"normalized": false,
|
135 |
+
"rstrip": true,
|
136 |
+
"single_word": false,
|
137 |
+
"special": true
|
138 |
+
},
|
139 |
+
"57534": {
|
140 |
+
"content": "<s_bill>",
|
141 |
+
"lstrip": true,
|
142 |
+
"normalized": false,
|
143 |
+
"rstrip": true,
|
144 |
+
"single_word": false,
|
145 |
+
"special": true
|
146 |
+
},
|
147 |
+
"57535": {
|
148 |
+
"content": "<s_description>",
|
149 |
+
"lstrip": true,
|
150 |
+
"normalized": false,
|
151 |
+
"rstrip": true,
|
152 |
+
"single_word": false,
|
153 |
+
"special": true
|
154 |
+
},
|
155 |
+
"57536": {
|
156 |
+
"content": "<s_address>",
|
157 |
+
"lstrip": true,
|
158 |
+
"normalized": false,
|
159 |
+
"rstrip": true,
|
160 |
+
"single_word": false,
|
161 |
+
"special": true
|
162 |
+
},
|
163 |
+
"57537": {
|
164 |
+
"content": "</s_item no>",
|
165 |
+
"lstrip": true,
|
166 |
+
"normalized": false,
|
167 |
+
"rstrip": true,
|
168 |
+
"single_word": false,
|
169 |
+
"special": true
|
170 |
+
},
|
171 |
+
"57538": {
|
172 |
+
"content": "<s_total>",
|
173 |
+
"lstrip": true,
|
174 |
+
"normalized": false,
|
175 |
+
"rstrip": true,
|
176 |
+
"single_word": false,
|
177 |
+
"special": true
|
178 |
+
},
|
179 |
+
"57539": {
|
180 |
+
"content": "<s_company>",
|
181 |
+
"lstrip": true,
|
182 |
+
"normalized": false,
|
183 |
+
"rstrip": true,
|
184 |
+
"single_word": false,
|
185 |
+
"special": true
|
186 |
+
},
|
187 |
+
"57540": {
|
188 |
+
"content": "</s_description>",
|
189 |
+
"lstrip": true,
|
190 |
+
"normalized": false,
|
191 |
+
"rstrip": true,
|
192 |
+
"single_word": false,
|
193 |
+
"special": true
|
194 |
}
|
195 |
},
|
196 |
"additional_special_tokens": [
|
197 |
"<s_iitcdip>",
|
198 |
"<s_synthdog>",
|
199 |
+
"<s_date>",
|
200 |
+
"</s_bill>",
|
201 |
+
"</s_date>",
|
202 |
+
"</s_address>",
|
203 |
+
"</s_total amount>",
|
204 |
+
"</s_total>",
|
205 |
+
"<s_total amount>",
|
206 |
+
"<s_item no>",
|
207 |
+
"</s_company>",
|
208 |
+
"<s_bill>",
|
209 |
+
"<s_description>",
|
210 |
+
"<s_address>",
|
211 |
+
"</s_item no>",
|
212 |
+
"<s_total>",
|
213 |
+
"<s_company>",
|
214 |
+
"</s_description>"
|
215 |
],
|
216 |
"bos_token": "<s>",
|
217 |
"clean_up_tokenization_spaces": true,
|