Hugging Face's logo

OpenVINO
/
Phi-3-mini-128k-instruct-int8-ov

Text Generation
Transformers
OpenVINO
phi3
conversational
custom_code
Inference Endpoints
Model card Files Community
katuni4ka commited on
Commit
41479e8
•
1 Parent(s): 1a1be1f

Upload 9 files

Browse files
Files changed (6) hide show
  1. openvino_detokenizer.bin +1 -1
  2. openvino_detokenizer.xml +10 -6
  3. openvino_tokenizer.bin +2 -2
  4. openvino_tokenizer.xml +246 -40
  5. tokenizer.json +1 -29
  6. tokenizer_config.json +4 -3
openvino_detokenizer.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:34f54bcfdecf199ed7333d7748c8e357c9d98b26c0c5800c7e0809e48edd74d3
3
  size 499991
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:630b6806812464da49d8dc0907d303055c3fa69f10b1f3533f6945437ab55b59
3
  size 499991
openvino_detokenizer.xml CHANGED
@@ -1,10 +1,10 @@
1
  <?xml version="1.0"?>
2
  <net name="detokenizer" version="11">
3
  <layers>
4
- <layer id="0" name="Parameter_22" type="Parameter" version="opset1">
5
  <data shape="?,?" element_type="i64" />
6
  <output>
7
- <port id="0" precision="I64" names="Parameter_22">
8
  <dim>-1</dim>
9
  <dim>-1</dim>
10
  </port>
@@ -18,7 +18,7 @@
18
  </port>
19
  </output>
20
  </layer>
21
- <layer id="2" name="Convert_32" type="Convert" version="opset1">
22
  <data destination_type="i32" />
23
  <input>
24
  <port id="0" precision="I64">
@@ -33,7 +33,7 @@
33
  </port>
34
  </output>
35
  </layer>
36
- <layer id="3" name="SentencepieceDetokenizer_23" type="SentencepieceDetokenizer" version="extension">
37
  <input>
38
  <port id="0" precision="U8">
39
  <dim>499991</dim>
@@ -55,7 +55,7 @@
55
  </port>
56
  </output>
57
  </layer>
58
- <layer id="4" name="StringTensorPack_24" type="StringTensorPack" version="extension">
59
  <data mode="begins_ends" />
60
  <input>
61
  <port id="0" precision="I32">
@@ -74,7 +74,7 @@
74
  </port>
75
  </output>
76
  </layer>
77
- <layer id="5" name="Result_25" type="Result" version="opset1">
78
  <input>
79
  <port id="0" precision="STRING">
80
  <dim>-1</dim>
@@ -92,6 +92,10 @@
92
  <edge from-layer="4" from-port="3" to-layer="5" to-port="0" />
93
  </edges>
94
  <rt_info>
 
 
95
  <eos_token_id value="32000" />
 
 
96
  </rt_info>
97
  </net>
 
1
  <?xml version="1.0"?>
2
  <net name="detokenizer" version="11">
3
  <layers>
4
+ <layer id="0" name="Parameter_37" type="Parameter" version="opset1">
5
  <data shape="?,?" element_type="i64" />
6
  <output>
7
+ <port id="0" precision="I64" names="Parameter_37">
8
  <dim>-1</dim>
9
  <dim>-1</dim>
10
  </port>
 
18
  </port>
19
  </output>
20
  </layer>
21
+ <layer id="2" name="Convert_47" type="Convert" version="opset1">
22
  <data destination_type="i32" />
23
  <input>
24
  <port id="0" precision="I64">
 
33
  </port>
34
  </output>
35
  </layer>
36
+ <layer id="3" name="SentencepieceDetokenizer_38" type="SentencepieceDetokenizer" version="extension">
37
  <input>
38
  <port id="0" precision="U8">
39
  <dim>499991</dim>
 
55
  </port>
56
  </output>
57
  </layer>
58
+ <layer id="4" name="StringTensorPack_39" type="StringTensorPack" version="extension">
59
  <data mode="begins_ends" />
60
  <input>
61
  <port id="0" precision="I32">
 
74
  </port>
75
  </output>
76
  </layer>
77
+ <layer id="5" name="Result_40" type="Result" version="opset1">
78
  <input>
79
  <port id="0" precision="STRING">
80
  <dim>-1</dim>
 
92
  <edge from-layer="4" from-port="3" to-layer="5" to-port="0" />
93
  </edges>
94
  <rt_info>
95
+ <bos_token_id value="1" />
96
+ <chat_template value="{% for message in messages %}{% if message['role'] == 'system' %}{{'&lt;|system|>&#10;' + message['content'] + '&lt;|end|>&#10;'}}{% elif message['role'] == 'user' %}{{'&lt;|user|>&#10;' + message['content'] + '&lt;|end|>&#10;'}}{% elif message['role'] == 'assistant' %}{{'&lt;|assistant|>&#10;' + message['content'] + '&lt;|end|>&#10;'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '&lt;|assistant|>&#10;' }}{% else %}{{ eos_token }}{% endif %}" />
97
  <eos_token_id value="32000" />
98
+ <original_tokenizer_class value="&lt;class 'transformers.models.llama.tokenization_llama_fast.LlamaTokenizerFast'>" />
99
+ <pad_token_id value="32000" />
100
  </rt_info>
101
  </net>
openvino_tokenizer.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:46097e0534935f1aec4cbac2c90e565ec51a8513fcd53b841231849403e5e122
3
- size 499981
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:996b60b8de0ee7597bb22dadabc6bed436a4255e0b2e2cbc7a11ad7ffbad8613
3
+ size 500278
openvino_tokenizer.xml CHANGED
@@ -9,7 +9,7 @@
9
  </port>
10
  </output>
11
  </layer>
12
- <layer id="1" name="Constant_8" type="Const" version="opset1">
13
  <data element_type="i32" shape="" offset="0" size="4" />
14
  <output>
15
  <port id="0" precision="I32" />
@@ -23,8 +23,128 @@
23
  </port>
24
  </output>
25
  </layer>
26
- <layer id="3" name="SentencepieceTokenizer_4" type="SentencepieceTokenizer" version="extension">
27
- <data nbest_size="0" alpha="0" add_bos="true" add_eos="false" reverse="false" />
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
  <input>
29
  <port id="0" precision="U8">
30
  <dim>499969</dim>
@@ -32,21 +152,33 @@
32
  <port id="1" precision="STRING">
33
  <dim>-1</dim>
34
  </port>
 
 
 
 
 
 
 
 
 
 
 
 
35
  </input>
36
  <output>
37
- <port id="2" precision="I64">
38
  <dim>-1</dim>
39
  <dim>2</dim>
40
  </port>
41
- <port id="3" precision="I32">
42
  <dim>-1</dim>
43
  </port>
44
- <port id="4" precision="I64">
45
  <dim>2</dim>
46
  </port>
47
  </output>
48
  </layer>
49
- <layer id="4" name="Broadcast_9" type="Broadcast" version="opset3">
50
  <data mode="numpy" />
51
  <input>
52
  <port id="0" precision="I32" />
@@ -61,13 +193,13 @@
61
  </port>
62
  </output>
63
  </layer>
64
- <layer id="5" name="Constant_10" type="Const" version="opset1">
65
- <data element_type="i32" shape="" offset="499973" size="4" />
66
  <output>
67
  <port id="0" precision="I32" />
68
  </output>
69
  </layer>
70
- <layer id="6" name="ShapeOf_11" type="ShapeOf" version="opset3">
71
  <data output_type="i64" />
72
  <input>
73
  <port id="0" precision="I32">
@@ -80,7 +212,7 @@
80
  </port>
81
  </output>
82
  </layer>
83
- <layer id="7" name="Broadcast_12" type="Broadcast" version="opset3">
84
  <data mode="numpy" />
85
  <input>
86
  <port id="0" precision="I32" />
@@ -94,7 +226,7 @@
94
  </port>
95
  </output>
96
  </layer>
97
- <layer id="8" name="ScatterNDUpdate_16" type="ScatterNDUpdate" version="opset4">
98
  <input>
99
  <port id="0" precision="I32">
100
  <dim>-1</dim>
@@ -115,7 +247,33 @@
115
  </port>
116
  </output>
117
  </layer>
118
- <layer id="9" name="ScatterNDUpdate_16" type="Convert" version="opset1">
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
119
  <data destination_type="i64" />
120
  <input>
121
  <port id="0" precision="I32">
@@ -130,14 +288,14 @@
130
  </port>
131
  </output>
132
  </layer>
133
- <layer id="11" name="Constant_5" type="Const" version="opset1">
134
- <data element_type="i32" shape="" offset="499977" size="4" />
135
  <output>
136
  <port id="0" precision="I32" />
137
  </output>
138
  </layer>
139
- <layer id="12" name="Broadcast_6" type="Broadcast" version="opset3">
140
- <data mode="numpy" />
141
  <input>
142
  <port id="0" precision="I32" />
143
  <port id="1" precision="I64">
@@ -151,7 +309,7 @@
151
  </port>
152
  </output>
153
  </layer>
154
- <layer id="13" name="ScatterNDUpdate_7" type="ScatterNDUpdate" version="opset4">
155
  <input>
156
  <port id="0" precision="I32">
157
  <dim>-1</dim>
@@ -172,7 +330,33 @@
172
  </port>
173
  </output>
174
  </layer>
175
- <layer id="14" name="ScatterNDUpdate_7" type="Convert" version="opset1">
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
176
  <data destination_type="i64" />
177
  <input>
178
  <port id="0" precision="I32">
@@ -187,7 +371,7 @@
187
  </port>
188
  </output>
189
  </layer>
190
- <layer id="15" name="Result_17" type="Result" version="opset1">
191
  <input>
192
  <port id="0" precision="I64">
193
  <dim>-1</dim>
@@ -195,7 +379,7 @@
195
  </port>
196
  </input>
197
  </layer>
198
- <layer id="10" name="Result_18" type="Result" version="opset1">
199
  <input>
200
  <port id="0" precision="I64">
201
  <dim>-1</dim>
@@ -205,27 +389,49 @@
205
  </layer>
206
  </layers>
207
  <edges>
208
- <edge from-layer="0" from-port="0" to-layer="3" to-port="1" />
209
- <edge from-layer="1" from-port="0" to-layer="4" to-port="0" />
210
- <edge from-layer="2" from-port="0" to-layer="3" to-port="0" />
211
- <edge from-layer="3" from-port="4" to-layer="4" to-port="1" />
212
- <edge from-layer="3" from-port="3" to-layer="6" to-port="0" />
213
- <edge from-layer="3" from-port="2" to-layer="8" to-port="1" />
214
- <edge from-layer="3" from-port="3" to-layer="13" to-port="2" />
215
- <edge from-layer="3" from-port="2" to-layer="13" to-port="1" />
216
- <edge from-layer="3" from-port="4" to-layer="12" to-port="1" />
217
- <edge from-layer="4" from-port="2" to-layer="8" to-port="0" />
218
- <edge from-layer="5" from-port="0" to-layer="7" to-port="0" />
219
- <edge from-layer="6" from-port="1" to-layer="7" to-port="1" />
220
- <edge from-layer="7" from-port="2" to-layer="8" to-port="2" />
221
- <edge from-layer="8" from-port="3" to-layer="9" to-port="0" />
222
- <edge from-layer="9" from-port="1" to-layer="10" to-port="0" />
223
- <edge from-layer="11" from-port="0" to-layer="12" to-port="0" />
224
- <edge from-layer="12" from-port="2" to-layer="13" to-port="0" />
225
- <edge from-layer="13" from-port="3" to-layer="14" to-port="0" />
226
- <edge from-layer="14" from-port="1" to-layer="15" to-port="0" />
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
227
  </edges>
228
  <rt_info>
 
 
229
  <eos_token_id value="32000" />
 
 
230
  </rt_info>
231
  </net>
 
9
  </port>
10
  </output>
11
  </layer>
12
+ <layer id="1" name="Constant_16" type="Const" version="opset1">
13
  <data element_type="i32" shape="" offset="0" size="4" />
14
  <output>
15
  <port id="0" precision="I32" />
 
23
  </port>
24
  </output>
25
  </layer>
26
+ <layer id="3" name="StringTensorUnpack_4" type="StringTensorUnpack" version="extension">
27
+ <data mode="begins_ends" />
28
+ <input>
29
+ <port id="0" precision="STRING">
30
+ <dim>-1</dim>
31
+ </port>
32
+ </input>
33
+ <output>
34
+ <port id="1" precision="I32">
35
+ <dim>-1</dim>
36
+ </port>
37
+ <port id="2" precision="I32">
38
+ <dim>-1</dim>
39
+ </port>
40
+ <port id="3" precision="U8">
41
+ <dim>-1</dim>
42
+ </port>
43
+ </output>
44
+ </layer>
45
+ <layer id="4" name="Constant_6" type="Const" version="opset1">
46
+ <data element_type="u8" shape="7" offset="499973" size="7" />
47
+ <output>
48
+ <port id="0" precision="U8">
49
+ <dim>7</dim>
50
+ </port>
51
+ </output>
52
+ </layer>
53
+ <layer id="5" name="Constant_8" type="Const" version="opset1">
54
+ <data element_type="u8" shape="3" offset="499980" size="3" />
55
+ <output>
56
+ <port id="0" precision="U8">
57
+ <dim>3</dim>
58
+ </port>
59
+ </output>
60
+ </layer>
61
+ <layer id="6" name="RegexNormalization_9" type="RegexNormalization" version="extension">
62
+ <data global_replace="true" />
63
+ <input>
64
+ <port id="0" precision="I32">
65
+ <dim>-1</dim>
66
+ </port>
67
+ <port id="1" precision="I32">
68
+ <dim>-1</dim>
69
+ </port>
70
+ <port id="2" precision="U8">
71
+ <dim>-1</dim>
72
+ </port>
73
+ <port id="3" precision="U8">
74
+ <dim>7</dim>
75
+ </port>
76
+ <port id="4" precision="U8">
77
+ <dim>3</dim>
78
+ </port>
79
+ </input>
80
+ <output>
81
+ <port id="5" precision="I32">
82
+ <dim>-1</dim>
83
+ </port>
84
+ <port id="6" precision="I32">
85
+ <dim>-1</dim>
86
+ </port>
87
+ <port id="7" precision="U8">
88
+ <dim>-1</dim>
89
+ </port>
90
+ </output>
91
+ </layer>
92
+ <layer id="7" name="StringTensorPack_10" type="StringTensorPack" version="extension">
93
+ <data mode="begins_ends" />
94
+ <input>
95
+ <port id="0" precision="I32">
96
+ <dim>-1</dim>
97
+ </port>
98
+ <port id="1" precision="I32">
99
+ <dim>-1</dim>
100
+ </port>
101
+ <port id="2" precision="U8">
102
+ <dim>-1</dim>
103
+ </port>
104
+ </input>
105
+ <output>
106
+ <port id="3" precision="STRING">
107
+ <dim>-1</dim>
108
+ </port>
109
+ </output>
110
+ </layer>
111
+ <layer id="8" name="Constant_12" type="Const" version="opset1">
112
+ <data element_type="u8" shape="223" offset="499983" size="223" />
113
+ <output>
114
+ <port id="0" precision="U8">
115
+ <dim>223</dim>
116
+ </port>
117
+ </output>
118
+ </layer>
119
+ <layer id="9" name="StringTensorUnpack_13" type="StringTensorUnpack" version="extension">
120
+ <data mode="begins_ends" />
121
+ <input>
122
+ <port id="0" precision="U8">
123
+ <dim>223</dim>
124
+ </port>
125
+ </input>
126
+ <output>
127
+ <port id="1" precision="I32">
128
+ <dim>-1</dim>
129
+ </port>
130
+ <port id="2" precision="I32">
131
+ <dim>-1</dim>
132
+ </port>
133
+ <port id="3" precision="U8">
134
+ <dim>-1</dim>
135
+ </port>
136
+ </output>
137
+ </layer>
138
+ <layer id="10" name="Constant_14" type="Const" version="opset1">
139
+ <data element_type="i32" shape="14" offset="500206" size="56" />
140
+ <output>
141
+ <port id="0" precision="I32">
142
+ <dim>14</dim>
143
+ </port>
144
+ </output>
145
+ </layer>
146
+ <layer id="11" name="SentencepieceTokenizer_15" type="SentencepieceTokenizer" version="extension">
147
+ <data nbest_size="0" alpha="0" add_bos="false" add_eos="false" reverse="true" />
148
  <input>
149
  <port id="0" precision="U8">
150
  <dim>499969</dim>
 
152
  <port id="1" precision="STRING">
153
  <dim>-1</dim>
154
  </port>
155
+ <port id="2" precision="I32">
156
+ <dim>-1</dim>
157
+ </port>
158
+ <port id="3" precision="I32">
159
+ <dim>-1</dim>
160
+ </port>
161
+ <port id="4" precision="U8">
162
+ <dim>-1</dim>
163
+ </port>
164
+ <port id="5" precision="I32">
165
+ <dim>14</dim>
166
+ </port>
167
  </input>
168
  <output>
169
+ <port id="6" precision="I64">
170
  <dim>-1</dim>
171
  <dim>2</dim>
172
  </port>
173
+ <port id="7" precision="I32">
174
  <dim>-1</dim>
175
  </port>
176
+ <port id="8" precision="I64">
177
  <dim>2</dim>
178
  </port>
179
  </output>
180
  </layer>
181
+ <layer id="12" name="Broadcast_17" type="Broadcast" version="opset3">
182
  <data mode="numpy" />
183
  <input>
184
  <port id="0" precision="I32" />
 
193
  </port>
194
  </output>
195
  </layer>
196
+ <layer id="13" name="Constant_18" type="Const" version="opset1">
197
+ <data element_type="i32" shape="" offset="500262" size="4" />
198
  <output>
199
  <port id="0" precision="I32" />
200
  </output>
201
  </layer>
202
+ <layer id="14" name="ShapeOf_19" type="ShapeOf" version="opset3">
203
  <data output_type="i64" />
204
  <input>
205
  <port id="0" precision="I32">
 
212
  </port>
213
  </output>
214
  </layer>
215
+ <layer id="15" name="Broadcast_20" type="Broadcast" version="opset3">
216
  <data mode="numpy" />
217
  <input>
218
  <port id="0" precision="I32" />
 
226
  </port>
227
  </output>
228
  </layer>
229
+ <layer id="16" name="ScatterNDUpdate_24" type="ScatterNDUpdate" version="opset4">
230
  <input>
231
  <port id="0" precision="I32">
232
  <dim>-1</dim>
 
247
  </port>
248
  </output>
249
  </layer>
250
+ <layer id="17" name="Constant_28" type="Const" version="opset1">
251
+ <data element_type="i64" shape="1" offset="500266" size="8" />
252
+ <output>
253
+ <port id="0" precision="I64">
254
+ <dim>1</dim>
255
+ </port>
256
+ </output>
257
+ </layer>
258
+ <layer id="18" name="Reverse_29" type="Reverse" version="opset1">
259
+ <data mode="index" />
260
+ <input>
261
+ <port id="0" precision="I32">
262
+ <dim>-1</dim>
263
+ <dim>-1</dim>
264
+ </port>
265
+ <port id="1" precision="I64">
266
+ <dim>1</dim>
267
+ </port>
268
+ </input>
269
+ <output>
270
+ <port id="2" precision="I32">
271
+ <dim>-1</dim>
272
+ <dim>-1</dim>
273
+ </port>
274
+ </output>
275
+ </layer>
276
+ <layer id="19" name="Reverse_29" type="Convert" version="opset1">
277
  <data destination_type="i64" />
278
  <input>
279
  <port id="0" precision="I32">
 
288
  </port>
289
  </output>
290
  </layer>
291
+ <layer id="21" name="Constant_25" type="Const" version="opset1">
292
+ <data element_type="i32" shape="" offset="500274" size="4" />
293
  <output>
294
  <port id="0" precision="I32" />
295
  </output>
296
  </layer>
297
+ <layer id="22" name="Broadcast_26" type="Broadcast" version="opset3">
298
+ <data mode="bidirectional" />
299
  <input>
300
  <port id="0" precision="I32" />
301
  <port id="1" precision="I64">
 
309
  </port>
310
  </output>
311
  </layer>
312
+ <layer id="23" name="ScatterNDUpdate_27" type="ScatterNDUpdate" version="opset4">
313
  <input>
314
  <port id="0" precision="I32">
315
  <dim>-1</dim>
 
330
  </port>
331
  </output>
332
  </layer>
333
+ <layer id="24" name="Constant_30" type="Const" version="opset1">
334
+ <data element_type="i64" shape="1" offset="500266" size="8" />
335
+ <output>
336
+ <port id="0" precision="I64">
337
+ <dim>1</dim>
338
+ </port>
339
+ </output>
340
+ </layer>
341
+ <layer id="25" name="Reverse_31" type="Reverse" version="opset1">
342
+ <data mode="index" />
343
+ <input>
344
+ <port id="0" precision="I32">
345
+ <dim>-1</dim>
346
+ <dim>-1</dim>
347
+ </port>
348
+ <port id="1" precision="I64">
349
+ <dim>1</dim>
350
+ </port>
351
+ </input>
352
+ <output>
353
+ <port id="2" precision="I32">
354
+ <dim>-1</dim>
355
+ <dim>-1</dim>
356
+ </port>
357
+ </output>
358
+ </layer>
359
+ <layer id="26" name="Reverse_31" type="Convert" version="opset1">
360
  <data destination_type="i64" />
361
  <input>
362
  <port id="0" precision="I32">
 
371
  </port>
372
  </output>
373
  </layer>
374
+ <layer id="27" name="Result_32" type="Result" version="opset1">
375
  <input>
376
  <port id="0" precision="I64">
377
  <dim>-1</dim>
 
379
  </port>
380
  </input>
381
  </layer>
382
+ <layer id="20" name="Result_33" type="Result" version="opset1">
383
  <input>
384
  <port id="0" precision="I64">
385
  <dim>-1</dim>
 
389
  </layer>
390
  </layers>
391
  <edges>
392
+ <edge from-layer="0" from-port="0" to-layer="3" to-port="0" />
393
+ <edge from-layer="1" from-port="0" to-layer="12" to-port="0" />
394
+ <edge from-layer="2" from-port="0" to-layer="11" to-port="0" />
395
+ <edge from-layer="3" from-port="1" to-layer="6" to-port="0" />
396
+ <edge from-layer="3" from-port="2" to-layer="6" to-port="1" />
397
+ <edge from-layer="3" from-port="3" to-layer="6" to-port="2" />
398
+ <edge from-layer="4" from-port="0" to-layer="6" to-port="3" />
399
+ <edge from-layer="5" from-port="0" to-layer="6" to-port="4" />
400
+ <edge from-layer="6" from-port="5" to-layer="7" to-port="0" />
401
+ <edge from-layer="6" from-port="6" to-layer="7" to-port="1" />
402
+ <edge from-layer="6" from-port="7" to-layer="7" to-port="2" />
403
+ <edge from-layer="7" from-port="3" to-layer="11" to-port="1" />
404
+ <edge from-layer="8" from-port="0" to-layer="9" to-port="0" />
405
+ <edge from-layer="9" from-port="1" to-layer="11" to-port="2" />
406
+ <edge from-layer="9" from-port="2" to-layer="11" to-port="3" />
407
+ <edge from-layer="9" from-port="3" to-layer="11" to-port="4" />
408
+ <edge from-layer="10" from-port="0" to-layer="11" to-port="5" />
409
+ <edge from-layer="11" from-port="7" to-layer="14" to-port="0" />
410
+ <edge from-layer="11" from-port="7" to-layer="23" to-port="2" />
411
+ <edge from-layer="11" from-port="6" to-layer="23" to-port="1" />
412
+ <edge from-layer="11" from-port="8" to-layer="22" to-port="1" />
413
+ <edge from-layer="11" from-port="6" to-layer="16" to-port="1" />
414
+ <edge from-layer="11" from-port="8" to-layer="12" to-port="1" />
415
+ <edge from-layer="12" from-port="2" to-layer="16" to-port="0" />
416
+ <edge from-layer="13" from-port="0" to-layer="15" to-port="0" />
417
+ <edge from-layer="14" from-port="1" to-layer="15" to-port="1" />
418
+ <edge from-layer="15" from-port="2" to-layer="16" to-port="2" />
419
+ <edge from-layer="16" from-port="3" to-layer="18" to-port="0" />
420
+ <edge from-layer="17" from-port="0" to-layer="18" to-port="1" />
421
+ <edge from-layer="18" from-port="2" to-layer="19" to-port="0" />
422
+ <edge from-layer="19" from-port="1" to-layer="20" to-port="0" />
423
+ <edge from-layer="21" from-port="0" to-layer="22" to-port="0" />
424
+ <edge from-layer="22" from-port="2" to-layer="23" to-port="0" />
425
+ <edge from-layer="23" from-port="3" to-layer="25" to-port="0" />
426
+ <edge from-layer="24" from-port="0" to-layer="25" to-port="1" />
427
+ <edge from-layer="25" from-port="2" to-layer="26" to-port="0" />
428
+ <edge from-layer="26" from-port="1" to-layer="27" to-port="0" />
429
  </edges>
430
  <rt_info>
431
+ <bos_token_id value="1" />
432
+ <chat_template value="{% for message in messages %}{% if message['role'] == 'system' %}{{'&lt;|system|>&#10;' + message['content'] + '&lt;|end|>&#10;'}}{% elif message['role'] == 'user' %}{{'&lt;|user|>&#10;' + message['content'] + '&lt;|end|>&#10;'}}{% elif message['role'] == 'assistant' %}{{'&lt;|assistant|>&#10;' + message['content'] + '&lt;|end|>&#10;'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '&lt;|assistant|>&#10;' }}{% else %}{{ eos_token }}{% endif %}" />
433
  <eos_token_id value="32000" />
434
+ <original_tokenizer_class value="&lt;class 'transformers.models.llama.tokenization_llama_fast.LlamaTokenizerFast'>" />
435
+ <pad_token_id value="32000" />
436
  </rt_info>
437
  </net>
tokenizer.json CHANGED
@@ -150,12 +150,6 @@
150
  "post_processor": {
151
  "type": "TemplateProcessing",
152
  "single": [
153
- {
154
- "SpecialToken": {
155
- "id": "<s>",
156
- "type_id": 0
157
- }
158
- },
159
  {
160
  "Sequence": {
161
  "id": "A",
@@ -164,24 +158,12 @@
164
  }
165
  ],
166
  "pair": [
167
- {
168
- "SpecialToken": {
169
- "id": "<s>",
170
- "type_id": 0
171
- }
172
- },
173
  {
174
  "Sequence": {
175
  "id": "A",
176
  "type_id": 0
177
  }
178
  },
179
- {
180
- "SpecialToken": {
181
- "id": "<s>",
182
- "type_id": 1
183
- }
184
- },
185
  {
186
  "Sequence": {
187
  "id": "B",
@@ -189,17 +171,7 @@
189
  }
190
  }
191
  ],
192
- "special_tokens": {
193
- "<s>": {
194
- "id": "<s>",
195
- "ids": [
196
- 1
197
- ],
198
- "tokens": [
199
- "<s>"
200
- ]
201
- }
202
- }
203
  },
204
  "decoder": {
205
  "type": "Sequence",
 
150
  "post_processor": {
151
  "type": "TemplateProcessing",
152
  "single": [
 
 
 
 
 
 
153
  {
154
  "Sequence": {
155
  "id": "A",
 
158
  }
159
  ],
160
  "pair": [
 
 
 
 
 
 
161
  {
162
  "Sequence": {
163
  "id": "A",
164
  "type_id": 0
165
  }
166
  },
 
 
 
 
 
 
167
  {
168
  "Sequence": {
169
  "id": "B",
 
171
  }
172
  }
173
  ],
174
+ "special_tokens": {}
 
 
 
 
 
 
 
 
 
 
175
  },
176
  "decoder": {
177
  "type": "Sequence",
tokenizer_config.json CHANGED
@@ -1,6 +1,7 @@
1
  {
2
- "add_bos_token": true,
3
  "add_eos_token": false,
 
4
  "added_tokens_decoder": {
5
  "0": {
6
  "content": "<unk>",
@@ -116,11 +117,11 @@
116
  }
117
  },
118
  "bos_token": "<s>",
119
- "chat_template": "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') %}{{'<|user|>' + '\n' + message['content'] + '<|end|>' + '\n' + '<|assistant|>' + '\n'}}{% elif (message['role'] == 'assistant') %}{{message['content'] + '<|end|>' + '\n'}}{% endif %}{% endfor %}",
120
  "clean_up_tokenization_spaces": false,
121
  "eos_token": "<|endoftext|>",
122
  "legacy": false,
123
- "model_max_length": 131072,
124
  "pad_token": "<|endoftext|>",
125
  "padding_side": "left",
126
  "sp_model_kwargs": {},
 
1
  {
2
+ "add_bos_token": false,
3
  "add_eos_token": false,
4
+ "add_prefix_space": null,
5
  "added_tokens_decoder": {
6
  "0": {
7
  "content": "<unk>",
 
117
  }
118
  },
119
  "bos_token": "<s>",
120
+ "chat_template": "{% for message in messages %}{% if message['role'] == 'system' %}{{'<|system|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'user' %}{{'<|user|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>\n' + message['content'] + '<|end|>\n'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}",
121
  "clean_up_tokenization_spaces": false,
122
  "eos_token": "<|endoftext|>",
123
  "legacy": false,
124
+ "model_max_length": 4096,
125
  "pad_token": "<|endoftext|>",
126
  "padding_side": "left",
127
  "sp_model_kwargs": {},