bofenghuang
commited on
Commit
·
1c47314
1
Parent(s):
1460c22
update vocab
Browse files- tokenizer.json +12 -12
- tokenizer_config.json +12 -12
tokenizer.json
CHANGED
@@ -104,7 +104,7 @@
|
|
104 |
},
|
105 |
{
|
106 |
"id": 128011,
|
107 |
-
"content": "<|
|
108 |
"single_word": false,
|
109 |
"lstrip": false,
|
110 |
"rstrip": false,
|
@@ -113,7 +113,7 @@
|
|
113 |
},
|
114 |
{
|
115 |
"id": 128012,
|
116 |
-
"content": "<|
|
117 |
"single_word": false,
|
118 |
"lstrip": false,
|
119 |
"rstrip": false,
|
@@ -122,7 +122,7 @@
|
|
122 |
},
|
123 |
{
|
124 |
"id": 128013,
|
125 |
-
"content": "<|
|
126 |
"single_word": false,
|
127 |
"lstrip": false,
|
128 |
"rstrip": false,
|
@@ -131,7 +131,7 @@
|
|
131 |
},
|
132 |
{
|
133 |
"id": 128014,
|
134 |
-
"content": "
|
135 |
"single_word": false,
|
136 |
"lstrip": false,
|
137 |
"rstrip": false,
|
@@ -140,7 +140,7 @@
|
|
140 |
},
|
141 |
{
|
142 |
"id": 128015,
|
143 |
-
"content": "
|
144 |
"single_word": false,
|
145 |
"lstrip": false,
|
146 |
"rstrip": false,
|
@@ -149,7 +149,7 @@
|
|
149 |
},
|
150 |
{
|
151 |
"id": 128016,
|
152 |
-
"content": "
|
153 |
"single_word": false,
|
154 |
"lstrip": false,
|
155 |
"rstrip": false,
|
@@ -158,7 +158,7 @@
|
|
158 |
},
|
159 |
{
|
160 |
"id": 128017,
|
161 |
-
"content": "
|
162 |
"single_word": false,
|
163 |
"lstrip": false,
|
164 |
"rstrip": false,
|
@@ -167,7 +167,7 @@
|
|
167 |
},
|
168 |
{
|
169 |
"id": 128018,
|
170 |
-
"content": "
|
171 |
"single_word": false,
|
172 |
"lstrip": false,
|
173 |
"rstrip": false,
|
@@ -176,7 +176,7 @@
|
|
176 |
},
|
177 |
{
|
178 |
"id": 128019,
|
179 |
-
"content": "
|
180 |
"single_word": false,
|
181 |
"lstrip": false,
|
182 |
"rstrip": false,
|
@@ -185,7 +185,7 @@
|
|
185 |
},
|
186 |
{
|
187 |
"id": 128020,
|
188 |
-
"content": "
|
189 |
"single_word": false,
|
190 |
"lstrip": false,
|
191 |
"rstrip": false,
|
@@ -194,7 +194,7 @@
|
|
194 |
},
|
195 |
{
|
196 |
"id": 128021,
|
197 |
-
"content": "
|
198 |
"single_word": false,
|
199 |
"lstrip": false,
|
200 |
"rstrip": false,
|
@@ -203,7 +203,7 @@
|
|
203 |
},
|
204 |
{
|
205 |
"id": 128022,
|
206 |
-
"content": "
|
207 |
"single_word": false,
|
208 |
"lstrip": false,
|
209 |
"rstrip": false,
|
|
|
104 |
},
|
105 |
{
|
106 |
"id": 128011,
|
107 |
+
"content": "<|system|>",
|
108 |
"single_word": false,
|
109 |
"lstrip": false,
|
110 |
"rstrip": false,
|
|
|
113 |
},
|
114 |
{
|
115 |
"id": 128012,
|
116 |
+
"content": "<|user|>",
|
117 |
"single_word": false,
|
118 |
"lstrip": false,
|
119 |
"rstrip": false,
|
|
|
122 |
},
|
123 |
{
|
124 |
"id": 128013,
|
125 |
+
"content": "<|assistant|>",
|
126 |
"single_word": false,
|
127 |
"lstrip": false,
|
128 |
"rstrip": false,
|
|
|
131 |
},
|
132 |
{
|
133 |
"id": 128014,
|
134 |
+
"content": "<available_tools>",
|
135 |
"single_word": false,
|
136 |
"lstrip": false,
|
137 |
"rstrip": false,
|
|
|
140 |
},
|
141 |
{
|
142 |
"id": 128015,
|
143 |
+
"content": "</available_tools>",
|
144 |
"single_word": false,
|
145 |
"lstrip": false,
|
146 |
"rstrip": false,
|
|
|
149 |
},
|
150 |
{
|
151 |
"id": 128016,
|
152 |
+
"content": "<tool_call>",
|
153 |
"single_word": false,
|
154 |
"lstrip": false,
|
155 |
"rstrip": false,
|
|
|
158 |
},
|
159 |
{
|
160 |
"id": 128017,
|
161 |
+
"content": "<tool_response>",
|
162 |
"single_word": false,
|
163 |
"lstrip": false,
|
164 |
"rstrip": false,
|
|
|
167 |
},
|
168 |
{
|
169 |
"id": 128018,
|
170 |
+
"content": "</tool_response>",
|
171 |
"single_word": false,
|
172 |
"lstrip": false,
|
173 |
"rstrip": false,
|
|
|
176 |
},
|
177 |
{
|
178 |
"id": 128019,
|
179 |
+
"content": "<available_documents>",
|
180 |
"single_word": false,
|
181 |
"lstrip": false,
|
182 |
"rstrip": false,
|
|
|
185 |
},
|
186 |
{
|
187 |
"id": 128020,
|
188 |
+
"content": "</available_documents>",
|
189 |
"single_word": false,
|
190 |
"lstrip": false,
|
191 |
"rstrip": false,
|
|
|
194 |
},
|
195 |
{
|
196 |
"id": 128021,
|
197 |
+
"content": "<reference>",
|
198 |
"single_word": false,
|
199 |
"lstrip": false,
|
200 |
"rstrip": false,
|
|
|
203 |
},
|
204 |
{
|
205 |
"id": 128022,
|
206 |
+
"content": "</reference>",
|
207 |
"single_word": false,
|
208 |
"lstrip": false,
|
209 |
"rstrip": false,
|
tokenizer_config.json
CHANGED
@@ -89,7 +89,7 @@
|
|
89 |
"special": true
|
90 |
},
|
91 |
"128011": {
|
92 |
-
"content": "<|
|
93 |
"lstrip": false,
|
94 |
"normalized": false,
|
95 |
"rstrip": false,
|
@@ -97,7 +97,7 @@
|
|
97 |
"special": true
|
98 |
},
|
99 |
"128012": {
|
100 |
-
"content": "<|
|
101 |
"lstrip": false,
|
102 |
"normalized": false,
|
103 |
"rstrip": false,
|
@@ -105,7 +105,7 @@
|
|
105 |
"special": true
|
106 |
},
|
107 |
"128013": {
|
108 |
-
"content": "<|
|
109 |
"lstrip": false,
|
110 |
"normalized": false,
|
111 |
"rstrip": false,
|
@@ -113,7 +113,7 @@
|
|
113 |
"special": true
|
114 |
},
|
115 |
"128014": {
|
116 |
-
"content": "
|
117 |
"lstrip": false,
|
118 |
"normalized": false,
|
119 |
"rstrip": false,
|
@@ -121,7 +121,7 @@
|
|
121 |
"special": true
|
122 |
},
|
123 |
"128015": {
|
124 |
-
"content": "
|
125 |
"lstrip": false,
|
126 |
"normalized": false,
|
127 |
"rstrip": false,
|
@@ -129,7 +129,7 @@
|
|
129 |
"special": true
|
130 |
},
|
131 |
"128016": {
|
132 |
-
"content": "
|
133 |
"lstrip": false,
|
134 |
"normalized": false,
|
135 |
"rstrip": false,
|
@@ -137,7 +137,7 @@
|
|
137 |
"special": true
|
138 |
},
|
139 |
"128017": {
|
140 |
-
"content": "
|
141 |
"lstrip": false,
|
142 |
"normalized": false,
|
143 |
"rstrip": false,
|
@@ -145,7 +145,7 @@
|
|
145 |
"special": true
|
146 |
},
|
147 |
"128018": {
|
148 |
-
"content": "
|
149 |
"lstrip": false,
|
150 |
"normalized": false,
|
151 |
"rstrip": false,
|
@@ -153,7 +153,7 @@
|
|
153 |
"special": true
|
154 |
},
|
155 |
"128019": {
|
156 |
-
"content": "
|
157 |
"lstrip": false,
|
158 |
"normalized": false,
|
159 |
"rstrip": false,
|
@@ -161,7 +161,7 @@
|
|
161 |
"special": true
|
162 |
},
|
163 |
"128020": {
|
164 |
-
"content": "
|
165 |
"lstrip": false,
|
166 |
"normalized": false,
|
167 |
"rstrip": false,
|
@@ -169,7 +169,7 @@
|
|
169 |
"special": true
|
170 |
},
|
171 |
"128021": {
|
172 |
-
"content": "
|
173 |
"lstrip": false,
|
174 |
"normalized": false,
|
175 |
"rstrip": false,
|
@@ -177,7 +177,7 @@
|
|
177 |
"special": true
|
178 |
},
|
179 |
"128022": {
|
180 |
-
"content": "
|
181 |
"lstrip": false,
|
182 |
"normalized": false,
|
183 |
"rstrip": false,
|
|
|
89 |
"special": true
|
90 |
},
|
91 |
"128011": {
|
92 |
+
"content": "<|system|>",
|
93 |
"lstrip": false,
|
94 |
"normalized": false,
|
95 |
"rstrip": false,
|
|
|
97 |
"special": true
|
98 |
},
|
99 |
"128012": {
|
100 |
+
"content": "<|user|>",
|
101 |
"lstrip": false,
|
102 |
"normalized": false,
|
103 |
"rstrip": false,
|
|
|
105 |
"special": true
|
106 |
},
|
107 |
"128013": {
|
108 |
+
"content": "<|assistant|>",
|
109 |
"lstrip": false,
|
110 |
"normalized": false,
|
111 |
"rstrip": false,
|
|
|
113 |
"special": true
|
114 |
},
|
115 |
"128014": {
|
116 |
+
"content": "<available_tools>",
|
117 |
"lstrip": false,
|
118 |
"normalized": false,
|
119 |
"rstrip": false,
|
|
|
121 |
"special": true
|
122 |
},
|
123 |
"128015": {
|
124 |
+
"content": "</available_tools>",
|
125 |
"lstrip": false,
|
126 |
"normalized": false,
|
127 |
"rstrip": false,
|
|
|
129 |
"special": true
|
130 |
},
|
131 |
"128016": {
|
132 |
+
"content": "<tool_call>",
|
133 |
"lstrip": false,
|
134 |
"normalized": false,
|
135 |
"rstrip": false,
|
|
|
137 |
"special": true
|
138 |
},
|
139 |
"128017": {
|
140 |
+
"content": "<tool_response>",
|
141 |
"lstrip": false,
|
142 |
"normalized": false,
|
143 |
"rstrip": false,
|
|
|
145 |
"special": true
|
146 |
},
|
147 |
"128018": {
|
148 |
+
"content": "</tool_response>",
|
149 |
"lstrip": false,
|
150 |
"normalized": false,
|
151 |
"rstrip": false,
|
|
|
153 |
"special": true
|
154 |
},
|
155 |
"128019": {
|
156 |
+
"content": "<available_documents>",
|
157 |
"lstrip": false,
|
158 |
"normalized": false,
|
159 |
"rstrip": false,
|
|
|
161 |
"special": true
|
162 |
},
|
163 |
"128020": {
|
164 |
+
"content": "</available_documents>",
|
165 |
"lstrip": false,
|
166 |
"normalized": false,
|
167 |
"rstrip": false,
|
|
|
169 |
"special": true
|
170 |
},
|
171 |
"128021": {
|
172 |
+
"content": "<reference>",
|
173 |
"lstrip": false,
|
174 |
"normalized": false,
|
175 |
"rstrip": false,
|
|
|
177 |
"special": true
|
178 |
},
|
179 |
"128022": {
|
180 |
+
"content": "</reference>",
|
181 |
"lstrip": false,
|
182 |
"normalized": false,
|
183 |
"rstrip": false,
|