Hugging Face's logo

FluidInference
/
phi-4-mini-instruct-int4-ov-npu

Text Generation
OpenVINO
phi3
nlp
code
conversational
custom_code
Model card Files
xet
Community
bweng commited on
Commit
089ceb0
·
verified ·
1 Parent(s): 1ffddaa

Upload folder using huggingface_hub

Browse files
Files changed (4) hide show
  1. openvino_detokenizer.xml +15 -17
  2. openvino_model.bin +2 -2
  3. openvino_model.xml +0 -0
  4. openvino_tokenizer.xml +49 -51
openvino_detokenizer.xml CHANGED
@@ -1,16 +1,16 @@
1
  <?xml version="1.0"?>
2
  <net name="detokenizer" version="11">
3
  <layers>
4
- <layer id="0" name="Parameter_184417" type="Parameter" version="opset1">
5
  <data shape="?,?" element_type="i64" />
6
  <output>
7
- <port id="0" precision="I64" names="Parameter_184417">
8
  <dim>-1</dim>
9
  <dim>-1</dim>
10
  </port>
11
  </output>
12
  </layer>
13
- <layer id="1" name="Convert_184607" type="Convert" version="opset1">
14
  <data destination_type="i32" />
15
  <input>
16
  <port id="0" precision="I64">
@@ -25,7 +25,7 @@
25
  </port>
26
  </output>
27
  </layer>
28
- <layer id="2" name="Constant_184419" type="Const" version="opset1">
29
  <data element_type="i32" shape="200029" offset="0" size="800116" />
30
  <output>
31
  <port id="0" precision="I32">
@@ -33,7 +33,7 @@
33
  </port>
34
  </output>
35
  </layer>
36
- <layer id="3" name="Constant_184421" type="Const" version="opset1">
37
  <data element_type="i32" shape="200029" offset="800116" size="800116" />
38
  <output>
39
  <port id="0" precision="I32">
@@ -41,7 +41,7 @@
41
  </port>
42
  </output>
43
  </layer>
44
- <layer id="4" name="Constant_184423" type="Const" version="opset1">
45
  <data element_type="u8" shape="1398089" offset="1600232" size="1398089" />
46
  <output>
47
  <port id="0" precision="U8">
@@ -49,7 +49,7 @@
49
  </port>
50
  </output>
51
  </layer>
52
- <layer id="5" name="Slice_184428" type="Const" version="opset1">
53
  <data element_type="i32" shape="7" offset="2998321" size="28" />
54
  <output>
55
  <port id="0" precision="I32">
@@ -57,7 +57,7 @@
57
  </port>
58
  </output>
59
  </layer>
60
- <layer id="6" name="VocabDecoder_184430" type="VocabDecoder" version="extension">
61
  <data skip_tokens="" />
62
  <input>
63
  <port id="0" precision="I32">
@@ -95,7 +95,7 @@
95
  </port>
96
  </output>
97
  </layer>
98
- <layer id="7" name="FuzeRagged_184431" type="FuzeRagged" version="extension">
99
  <input>
100
  <port id="0" precision="I32">
101
  <dim>-1</dim>
@@ -119,7 +119,7 @@
119
  </port>
120
  </output>
121
  </layer>
122
- <layer id="8" name="UTF8Validate_184432" type="UTF8Validate" version="extension">
123
  <data replace_mode="true" />
124
  <input>
125
  <port id="0" precision="I32">
@@ -144,7 +144,7 @@
144
  </port>
145
  </output>
146
  </layer>
147
- <layer id="9" name="StringTensorPack_184433" type="StringTensorPack" version="opset15">
148
  <input>
149
  <port id="0" precision="I32">
150
  <dim>-1</dim>
@@ -157,12 +157,12 @@
157
  </port>
158
  </input>
159
  <output>
160
- <port id="3" precision="STRING" names="Result_184434,string_output">
161
  <dim>-1</dim>
162
  </port>
163
  </output>
164
  </layer>
165
- <layer id="10" name="Result_184434" type="Result" version="opset1" output_names="Result_184434,string_output">
166
  <input>
167
  <port id="0" precision="STRING">
168
  <dim>-1</dim>
@@ -201,14 +201,12 @@
201
  <handle_special_tokens_with_re />
202
  <max_length />
203
  <number_of_inputs value="1" />
204
- <openvino_tokenizers_version value="2025.2.0.0-548-77f7abc0a90" />
205
- <openvino_version value="2025.2.0-18823-4ebe493140b" />
206
  <original_tokenizer_class value="&lt;class 'transformers.models.gpt2.tokenization_gpt2_fast.GPT2TokenizerFast'>" />
207
  <pad_token_id value="199999" />
208
- <sentencepiece_version value="0.2.0" />
209
  <skip_special_tokens value="True" />
210
  <streaming_detokenizer value="False" />
211
- <tiktoken_version value="0.9.0" />
212
  <tokenizer_output_type value="i64" />
213
  <tokenizers_version value="0.21.1" />
214
  <transformers_version value="4.51.3" />
 
1
  <?xml version="1.0"?>
2
  <net name="detokenizer" version="11">
3
  <layers>
4
+ <layer id="0" name="Parameter_184481" type="Parameter" version="opset1">
5
  <data shape="?,?" element_type="i64" />
6
  <output>
7
+ <port id="0" precision="I64" names="Parameter_184481">
8
  <dim>-1</dim>
9
  <dim>-1</dim>
10
  </port>
11
  </output>
12
  </layer>
13
+ <layer id="1" name="Convert_184671" type="Convert" version="opset1">
14
  <data destination_type="i32" />
15
  <input>
16
  <port id="0" precision="I64">
 
25
  </port>
26
  </output>
27
  </layer>
28
+ <layer id="2" name="Constant_184483" type="Const" version="opset1">
29
  <data element_type="i32" shape="200029" offset="0" size="800116" />
30
  <output>
31
  <port id="0" precision="I32">
 
33
  </port>
34
  </output>
35
  </layer>
36
+ <layer id="3" name="Constant_184485" type="Const" version="opset1">
37
  <data element_type="i32" shape="200029" offset="800116" size="800116" />
38
  <output>
39
  <port id="0" precision="I32">
 
41
  </port>
42
  </output>
43
  </layer>
44
+ <layer id="4" name="Constant_184487" type="Const" version="opset1">
45
  <data element_type="u8" shape="1398089" offset="1600232" size="1398089" />
46
  <output>
47
  <port id="0" precision="U8">
 
49
  </port>
50
  </output>
51
  </layer>
52
+ <layer id="5" name="Slice_184492" type="Const" version="opset1">
53
  <data element_type="i32" shape="7" offset="2998321" size="28" />
54
  <output>
55
  <port id="0" precision="I32">
 
57
  </port>
58
  </output>
59
  </layer>
60
+ <layer id="6" name="VocabDecoder_184494" type="VocabDecoder" version="extension">
61
  <data skip_tokens="" />
62
  <input>
63
  <port id="0" precision="I32">
 
95
  </port>
96
  </output>
97
  </layer>
98
+ <layer id="7" name="FuzeRagged_184495" type="FuzeRagged" version="extension">
99
  <input>
100
  <port id="0" precision="I32">
101
  <dim>-1</dim>
 
119
  </port>
120
  </output>
121
  </layer>
122
+ <layer id="8" name="UTF8Validate_184496" type="UTF8Validate" version="extension">
123
  <data replace_mode="true" />
124
  <input>
125
  <port id="0" precision="I32">
 
144
  </port>
145
  </output>
146
  </layer>
147
+ <layer id="9" name="StringTensorPack_184497" type="StringTensorPack" version="opset15">
148
  <input>
149
  <port id="0" precision="I32">
150
  <dim>-1</dim>
 
157
  </port>
158
  </input>
159
  <output>
160
+ <port id="3" precision="STRING" names="Result_184498,string_output">
161
  <dim>-1</dim>
162
  </port>
163
  </output>
164
  </layer>
165
+ <layer id="10" name="Result_184498" type="Result" version="opset1" output_names="Result_184498,string_output">
166
  <input>
167
  <port id="0" precision="STRING">
168
  <dim>-1</dim>
 
201
  <handle_special_tokens_with_re />
202
  <max_length />
203
  <number_of_inputs value="1" />
204
+ <openvino_tokenizers_version value="2025.2.0.0-566-85be884a69f" />
205
+ <openvino_version value="2025.2.0-19140-c01cd93e24d-releases/2025/2" />
206
  <original_tokenizer_class value="&lt;class 'transformers.models.gpt2.tokenization_gpt2_fast.GPT2TokenizerFast'>" />
207
  <pad_token_id value="199999" />
 
208
  <skip_special_tokens value="True" />
209
  <streaming_detokenizer value="False" />
 
210
  <tokenizer_output_type value="i64" />
211
  <tokenizers_version value="0.21.1" />
212
  <transformers_version value="4.51.3" />
openvino_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2172ddd59c274760a531c00042ce8605927bed72e164d9e554acc4f82b122163
3
- size 2228378220
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:394d65df4da3f975d383a3e1e9515cd2039c0dc3d508e93e8f54ac63c046183c
3
+ size 2228378216
openvino_model.xml CHANGED
The diff for this file is too large to render. See raw diff
 
openvino_tokenizer.xml CHANGED
@@ -1,21 +1,21 @@
1
  <?xml version="1.0"?>
2
  <net name="tokenizer" version="11">
3
  <layers>
4
- <layer id="0" name="Parameter_184297" type="Parameter" version="opset1">
5
  <data shape="?" element_type="string" />
6
  <output>
7
- <port id="0" precision="STRING" names="Parameter_184297">
8
  <dim>-1</dim>
9
  </port>
10
  </output>
11
  </layer>
12
- <layer id="1" name="Constant_184303" type="Const" version="opset1">
13
  <data element_type="i64" shape="" offset="0" size="8" />
14
  <output>
15
  <port id="0" precision="I64" />
16
  </output>
17
  </layer>
18
- <layer id="2" name="StringTensorUnpack_184298" type="StringTensorUnpack" version="opset15">
19
  <input>
20
  <port id="0" precision="STRING">
21
  <dim>-1</dim>
@@ -33,7 +33,7 @@
33
  </port>
34
  </output>
35
  </layer>
36
- <layer id="3" name="ShapeOf_184299" type="ShapeOf" version="opset3">
37
  <data output_type="i64" />
38
  <input>
39
  <port id="0" precision="I32">
@@ -46,19 +46,19 @@
46
  </port>
47
  </output>
48
  </layer>
49
- <layer id="4" name="Constant_184300" type="Const" version="opset1">
50
  <data element_type="i64" shape="" offset="0" size="8" />
51
  <output>
52
  <port id="0" precision="I64" />
53
  </output>
54
  </layer>
55
- <layer id="5" name="Constant_184301" type="Const" version="opset1">
56
  <data element_type="i64" shape="" offset="0" size="8" />
57
  <output>
58
  <port id="0" precision="I64" />
59
  </output>
60
  </layer>
61
- <layer id="6" name="Gather_184302" type="Gather" version="opset8">
62
  <data batch_dims="0" />
63
  <input>
64
  <port id="0" precision="I64">
@@ -71,13 +71,13 @@
71
  <port id="3" precision="I64" />
72
  </output>
73
  </layer>
74
- <layer id="7" name="Constant_184304" type="Const" version="opset1">
75
  <data element_type="i64" shape="" offset="8" size="8" />
76
  <output>
77
  <port id="0" precision="I64" />
78
  </output>
79
  </layer>
80
- <layer id="8" name="Range_184305" type="Range" version="opset4">
81
  <data output_type="i32" />
82
  <input>
83
  <port id="0" precision="I64" />
@@ -90,19 +90,19 @@
90
  </port>
91
  </output>
92
  </layer>
93
- <layer id="9" name="Constant_184306" type="Const" version="opset1">
94
  <data element_type="i64" shape="" offset="8" size="8" />
95
  <output>
96
  <port id="0" precision="I64" />
97
  </output>
98
  </layer>
99
- <layer id="10" name="Constant_184307" type="Const" version="opset1">
100
  <data element_type="i64" shape="" offset="8" size="8" />
101
  <output>
102
  <port id="0" precision="I64" />
103
  </output>
104
  </layer>
105
- <layer id="11" name="Add_184308" type="Add" version="opset1">
106
  <data auto_broadcast="numpy" />
107
  <input>
108
  <port id="0" precision="I64" />
@@ -112,13 +112,13 @@
112
  <port id="2" precision="I64" />
113
  </output>
114
  </layer>
115
- <layer id="12" name="Constant_184309" type="Const" version="opset1">
116
  <data element_type="i64" shape="" offset="8" size="8" />
117
  <output>
118
  <port id="0" precision="I64" />
119
  </output>
120
  </layer>
121
- <layer id="13" name="Range_184310" type="Range" version="opset4">
122
  <data output_type="i32" />
123
  <input>
124
  <port id="0" precision="I64" />
@@ -131,7 +131,7 @@
131
  </port>
132
  </output>
133
  </layer>
134
- <layer id="14" name="Constant_184372" type="Const" version="opset1">
135
  <data element_type="u8" shape="289" offset="16" size="289" />
136
  <output>
137
  <port id="0" precision="U8">
@@ -139,7 +139,7 @@
139
  </port>
140
  </output>
141
  </layer>
142
- <layer id="15" name="SpecialTokensSplit_184373" type="SpecialTokensSplit" version="extension">
143
  <input>
144
  <port id="0" precision="I32">
145
  <dim>-1</dim>
@@ -181,7 +181,7 @@
181
  </port>
182
  </output>
183
  </layer>
184
- <layer id="16" name="Constant_184375" type="Const" version="opset1">
185
  <data element_type="u8" shape="274" offset="305" size="274" />
186
  <output>
187
  <port id="0" precision="U8">
@@ -189,7 +189,7 @@
189
  </port>
190
  </output>
191
  </layer>
192
- <layer id="17" name="RegexSplit_184376" type="RegexSplit" version="extension">
193
  <data behaviour="remove" invert="true" max_splits="-1" />
194
  <input>
195
  <port id="0" precision="I32">
@@ -235,7 +235,7 @@
235
  </port>
236
  </output>
237
  </layer>
238
- <layer id="18" name="Constant_184378" type="Const" version="opset1">
239
  <data element_type="i32" shape="200029" offset="579" size="800116" />
240
  <output>
241
  <port id="0" precision="I32">
@@ -243,7 +243,7 @@
243
  </port>
244
  </output>
245
  </layer>
246
- <layer id="19" name="Constant_184380" type="Const" version="opset1">
247
  <data element_type="i32" shape="200029" offset="800695" size="800116" />
248
  <output>
249
  <port id="0" precision="I32">
@@ -251,7 +251,7 @@
251
  </port>
252
  </output>
253
  </layer>
254
- <layer id="20" name="Constant_184382" type="Const" version="opset1">
255
  <data element_type="u8" shape="1398109" offset="1600811" size="1398109" />
256
  <output>
257
  <port id="0" precision="U8">
@@ -259,7 +259,7 @@
259
  </port>
260
  </output>
261
  </layer>
262
- <layer id="21" name="Constant_184390" type="Const" version="opset1">
263
  <data element_type="i32" shape="199742" offset="2998920" size="798968" />
264
  <output>
265
  <port id="0" precision="I32">
@@ -267,7 +267,7 @@
267
  </port>
268
  </output>
269
  </layer>
270
- <layer id="22" name="Constant_184392" type="Const" version="opset1">
271
  <data element_type="i32" shape="199742" offset="3797888" size="798968" />
272
  <output>
273
  <port id="0" precision="I32">
@@ -275,7 +275,7 @@
275
  </port>
276
  </output>
277
  </layer>
278
- <layer id="23" name="Constant_184394" type="Const" version="opset1">
279
  <data element_type="u8" shape="718313" offset="4596856" size="718313" />
280
  <output>
281
  <port id="0" precision="U8">
@@ -283,7 +283,7 @@
283
  </port>
284
  </output>
285
  </layer>
286
- <layer id="24" name="Constant_184396" type="Const" version="opset1">
287
  <data element_type="i32" shape="199742" offset="5315169" size="798968" />
288
  <output>
289
  <port id="0" precision="I32">
@@ -291,7 +291,7 @@
291
  </port>
292
  </output>
293
  </layer>
294
- <layer id="25" name="Constant_184398" type="Const" version="opset1">
295
  <data element_type="i32" shape="199742" offset="6114137" size="798968" />
296
  <output>
297
  <port id="0" precision="I32">
@@ -299,7 +299,7 @@
299
  </port>
300
  </output>
301
  </layer>
302
- <layer id="26" name="Constant_184400" type="Const" version="opset1">
303
  <data element_type="u8" shape="679101" offset="6913105" size="679101" />
304
  <output>
305
  <port id="0" precision="U8">
@@ -307,7 +307,7 @@
307
  </port>
308
  </output>
309
  </layer>
310
- <layer id="27" name="Constant_184384" type="Const" version="opset1">
311
  <data element_type="i32" shape="432" offset="7592206" size="1728" />
312
  <output>
313
  <port id="0" precision="I32">
@@ -315,7 +315,7 @@
315
  </port>
316
  </output>
317
  </layer>
318
- <layer id="28" name="Constant_184386" type="Const" version="opset1">
319
  <data element_type="i32" shape="432" offset="7593934" size="1728" />
320
  <output>
321
  <port id="0" precision="I32">
@@ -323,7 +323,7 @@
323
  </port>
324
  </output>
325
  </layer>
326
- <layer id="29" name="Constant_184388" type="Const" version="opset1">
327
  <data element_type="u8" shape="5366" offset="7595662" size="5366" />
328
  <output>
329
  <port id="0" precision="U8">
@@ -331,7 +331,7 @@
331
  </port>
332
  </output>
333
  </layer>
334
- <layer id="30" name="Constant_184401" type="Const" version="opset1">
335
  <data element_type="i32" shape="432" offset="7601028" size="1728" />
336
  <output>
337
  <port id="0" precision="I32">
@@ -339,7 +339,7 @@
339
  </port>
340
  </output>
341
  </layer>
342
- <layer id="31" name="BPETokenizer_184402" type="BPETokenizer" version="extension">
343
  <data unk_token="" fuse_unk="false" suffix_indicator="" end_suffix="" byte_fallback="false" cache_capacity="40003" />
344
  <input>
345
  <port id="0" precision="I32">
@@ -409,13 +409,13 @@
409
  </port>
410
  </output>
411
  </layer>
412
- <layer id="32" name="Constant_184403" type="Const" version="opset1">
413
  <data element_type="i32" shape="" offset="7602756" size="4" />
414
  <output>
415
  <port id="0" precision="I32" />
416
  </output>
417
  </layer>
418
- <layer id="33" name="Constant_184405" type="Const" version="opset1">
419
  <data element_type="u8" shape="4" offset="7602760" size="4" />
420
  <output>
421
  <port id="0" precision="U8">
@@ -423,7 +423,7 @@
423
  </port>
424
  </output>
425
  </layer>
426
- <layer id="34" name="Constant_184407" type="Const" version="opset1">
427
  <data element_type="u8" shape="13" offset="7602764" size="13" />
428
  <output>
429
  <port id="0" precision="U8">
@@ -431,7 +431,7 @@
431
  </port>
432
  </output>
433
  </layer>
434
- <layer id="35" name="Truncate_184408" type="Truncate" version="extension">
435
  <data m_num_inputs="1" />
436
  <input>
437
  <port id="0" precision="I32">
@@ -463,7 +463,7 @@
463
  </port>
464
  </output>
465
  </layer>
466
- <layer id="36" name="Subtract_184409" type="Subtract" version="opset1">
467
  <data auto_broadcast="numpy" />
468
  <input>
469
  <port id="0" precision="I32">
@@ -479,13 +479,13 @@
479
  </port>
480
  </output>
481
  </layer>
482
- <layer id="37" name="Constant_184410" type="Const" version="opset1">
483
  <data element_type="i32" shape="" offset="7602777" size="4" />
484
  <output>
485
  <port id="0" precision="I32" />
486
  </output>
487
  </layer>
488
- <layer id="38" name="ReduceMax_184411" type="ReduceMax" version="opset1">
489
  <data keep_dims="false" />
490
  <input>
491
  <port id="0" precision="I32">
@@ -497,13 +497,13 @@
497
  <port id="2" precision="I32" />
498
  </output>
499
  </layer>
500
- <layer id="39" name="Constant_184412" type="Const" version="opset1">
501
  <data element_type="i32" shape="" offset="7602781" size="4" />
502
  <output>
503
  <port id="0" precision="I32" />
504
  </output>
505
  </layer>
506
- <layer id="40" name="RaggedToDense_184413" type="RaggedToDense" version="extension">
507
  <data pad_right="false" m_pad_max_length="false" />
508
  <input>
509
  <port id="0" precision="I32">
@@ -529,7 +529,7 @@
529
  </port>
530
  </output>
531
  </layer>
532
- <layer id="41" name="Convert_184414" type="Convert" version="opset1">
533
  <data destination_type="i32" />
534
  <input>
535
  <port id="0" precision="BOOL">
@@ -544,7 +544,7 @@
544
  </port>
545
  </output>
546
  </layer>
547
- <layer id="42" name="Convert_184414.0" type="Convert" version="opset1">
548
  <data destination_type="i64" />
549
  <input>
550
  <port id="0" precision="I32">
@@ -559,7 +559,7 @@
559
  </port>
560
  </output>
561
  </layer>
562
- <layer id="44" name="RaggedToDense_184413.0" type="Convert" version="opset1">
563
  <data destination_type="i64" />
564
  <input>
565
  <port id="0" precision="I32">
@@ -574,7 +574,7 @@
574
  </port>
575
  </output>
576
  </layer>
577
- <layer id="45" name="Result_184415" type="Result" version="opset1" output_names="input_ids">
578
  <input>
579
  <port id="0" precision="I64">
580
  <dim>-1</dim>
@@ -582,7 +582,7 @@
582
  </port>
583
  </input>
584
  </layer>
585
- <layer id="43" name="Result_184416" type="Result" version="opset1" output_names="attention_mask">
586
  <input>
587
  <port id="0" precision="I64">
588
  <dim>-1</dim>
@@ -669,14 +669,12 @@
669
  <handle_special_tokens_with_re />
670
  <max_length />
671
  <number_of_inputs value="1" />
672
- <openvino_tokenizers_version value="2025.2.0.0-548-77f7abc0a90" />
673
- <openvino_version value="2025.2.0-18823-4ebe493140b" />
674
  <original_tokenizer_class value="&lt;class 'transformers.models.gpt2.tokenization_gpt2_fast.GPT2TokenizerFast'>" />
675
  <pad_token_id value="199999" />
676
- <sentencepiece_version value="0.2.0" />
677
  <skip_special_tokens value="True" />
678
  <streaming_detokenizer value="False" />
679
- <tiktoken_version value="0.9.0" />
680
  <tokenizer_output_type value="i64" />
681
  <tokenizers_version value="0.21.1" />
682
  <transformers_version value="4.51.3" />
 
1
  <?xml version="1.0"?>
2
  <net name="tokenizer" version="11">
3
  <layers>
4
+ <layer id="0" name="Parameter_184361" type="Parameter" version="opset1">
5
  <data shape="?" element_type="string" />
6
  <output>
7
+ <port id="0" precision="STRING" names="Parameter_184361">
8
  <dim>-1</dim>
9
  </port>
10
  </output>
11
  </layer>
12
+ <layer id="1" name="Constant_184367" type="Const" version="opset1">
13
  <data element_type="i64" shape="" offset="0" size="8" />
14
  <output>
15
  <port id="0" precision="I64" />
16
  </output>
17
  </layer>
18
+ <layer id="2" name="StringTensorUnpack_184362" type="StringTensorUnpack" version="opset15">
19
  <input>
20
  <port id="0" precision="STRING">
21
  <dim>-1</dim>
 
33
  </port>
34
  </output>
35
  </layer>
36
+ <layer id="3" name="ShapeOf_184363" type="ShapeOf" version="opset3">
37
  <data output_type="i64" />
38
  <input>
39
  <port id="0" precision="I32">
 
46
  </port>
47
  </output>
48
  </layer>
49
+ <layer id="4" name="Constant_184364" type="Const" version="opset1">
50
  <data element_type="i64" shape="" offset="0" size="8" />
51
  <output>
52
  <port id="0" precision="I64" />
53
  </output>
54
  </layer>
55
+ <layer id="5" name="Constant_184365" type="Const" version="opset1">
56
  <data element_type="i64" shape="" offset="0" size="8" />
57
  <output>
58
  <port id="0" precision="I64" />
59
  </output>
60
  </layer>
61
+ <layer id="6" name="Gather_184366" type="Gather" version="opset8">
62
  <data batch_dims="0" />
63
  <input>
64
  <port id="0" precision="I64">
 
71
  <port id="3" precision="I64" />
72
  </output>
73
  </layer>
74
+ <layer id="7" name="Constant_184368" type="Const" version="opset1">
75
  <data element_type="i64" shape="" offset="8" size="8" />
76
  <output>
77
  <port id="0" precision="I64" />
78
  </output>
79
  </layer>
80
+ <layer id="8" name="Range_184369" type="Range" version="opset4">
81
  <data output_type="i32" />
82
  <input>
83
  <port id="0" precision="I64" />
 
90
  </port>
91
  </output>
92
  </layer>
93
+ <layer id="9" name="Constant_184370" type="Const" version="opset1">
94
  <data element_type="i64" shape="" offset="8" size="8" />
95
  <output>
96
  <port id="0" precision="I64" />
97
  </output>
98
  </layer>
99
+ <layer id="10" name="Constant_184371" type="Const" version="opset1">
100
  <data element_type="i64" shape="" offset="8" size="8" />
101
  <output>
102
  <port id="0" precision="I64" />
103
  </output>
104
  </layer>
105
+ <layer id="11" name="Add_184372" type="Add" version="opset1">
106
  <data auto_broadcast="numpy" />
107
  <input>
108
  <port id="0" precision="I64" />
 
112
  <port id="2" precision="I64" />
113
  </output>
114
  </layer>
115
+ <layer id="12" name="Constant_184373" type="Const" version="opset1">
116
  <data element_type="i64" shape="" offset="8" size="8" />
117
  <output>
118
  <port id="0" precision="I64" />
119
  </output>
120
  </layer>
121
+ <layer id="13" name="Range_184374" type="Range" version="opset4">
122
  <data output_type="i32" />
123
  <input>
124
  <port id="0" precision="I64" />
 
131
  </port>
132
  </output>
133
  </layer>
134
+ <layer id="14" name="Constant_184436" type="Const" version="opset1">
135
  <data element_type="u8" shape="289" offset="16" size="289" />
136
  <output>
137
  <port id="0" precision="U8">
 
139
  </port>
140
  </output>
141
  </layer>
142
+ <layer id="15" name="SpecialTokensSplit_184437" type="SpecialTokensSplit" version="extension">
143
  <input>
144
  <port id="0" precision="I32">
145
  <dim>-1</dim>
 
181
  </port>
182
  </output>
183
  </layer>
184
+ <layer id="16" name="Constant_184439" type="Const" version="opset1">
185
  <data element_type="u8" shape="274" offset="305" size="274" />
186
  <output>
187
  <port id="0" precision="U8">
 
189
  </port>
190
  </output>
191
  </layer>
192
+ <layer id="17" name="RegexSplit_184440" type="RegexSplit" version="extension">
193
  <data behaviour="remove" invert="true" max_splits="-1" />
194
  <input>
195
  <port id="0" precision="I32">
 
235
  </port>
236
  </output>
237
  </layer>
238
+ <layer id="18" name="Constant_184442" type="Const" version="opset1">
239
  <data element_type="i32" shape="200029" offset="579" size="800116" />
240
  <output>
241
  <port id="0" precision="I32">
 
243
  </port>
244
  </output>
245
  </layer>
246
+ <layer id="19" name="Constant_184444" type="Const" version="opset1">
247
  <data element_type="i32" shape="200029" offset="800695" size="800116" />
248
  <output>
249
  <port id="0" precision="I32">
 
251
  </port>
252
  </output>
253
  </layer>
254
+ <layer id="20" name="Constant_184446" type="Const" version="opset1">
255
  <data element_type="u8" shape="1398109" offset="1600811" size="1398109" />
256
  <output>
257
  <port id="0" precision="U8">
 
259
  </port>
260
  </output>
261
  </layer>
262
+ <layer id="21" name="Constant_184454" type="Const" version="opset1">
263
  <data element_type="i32" shape="199742" offset="2998920" size="798968" />
264
  <output>
265
  <port id="0" precision="I32">
 
267
  </port>
268
  </output>
269
  </layer>
270
+ <layer id="22" name="Constant_184456" type="Const" version="opset1">
271
  <data element_type="i32" shape="199742" offset="3797888" size="798968" />
272
  <output>
273
  <port id="0" precision="I32">
 
275
  </port>
276
  </output>
277
  </layer>
278
+ <layer id="23" name="Constant_184458" type="Const" version="opset1">
279
  <data element_type="u8" shape="718313" offset="4596856" size="718313" />
280
  <output>
281
  <port id="0" precision="U8">
 
283
  </port>
284
  </output>
285
  </layer>
286
+ <layer id="24" name="Constant_184460" type="Const" version="opset1">
287
  <data element_type="i32" shape="199742" offset="5315169" size="798968" />
288
  <output>
289
  <port id="0" precision="I32">
 
291
  </port>
292
  </output>
293
  </layer>
294
+ <layer id="25" name="Constant_184462" type="Const" version="opset1">
295
  <data element_type="i32" shape="199742" offset="6114137" size="798968" />
296
  <output>
297
  <port id="0" precision="I32">
 
299
  </port>
300
  </output>
301
  </layer>
302
+ <layer id="26" name="Constant_184464" type="Const" version="opset1">
303
  <data element_type="u8" shape="679101" offset="6913105" size="679101" />
304
  <output>
305
  <port id="0" precision="U8">
 
307
  </port>
308
  </output>
309
  </layer>
310
+ <layer id="27" name="Constant_184448" type="Const" version="opset1">
311
  <data element_type="i32" shape="432" offset="7592206" size="1728" />
312
  <output>
313
  <port id="0" precision="I32">
 
315
  </port>
316
  </output>
317
  </layer>
318
+ <layer id="28" name="Constant_184450" type="Const" version="opset1">
319
  <data element_type="i32" shape="432" offset="7593934" size="1728" />
320
  <output>
321
  <port id="0" precision="I32">
 
323
  </port>
324
  </output>
325
  </layer>
326
+ <layer id="29" name="Constant_184452" type="Const" version="opset1">
327
  <data element_type="u8" shape="5366" offset="7595662" size="5366" />
328
  <output>
329
  <port id="0" precision="U8">
 
331
  </port>
332
  </output>
333
  </layer>
334
+ <layer id="30" name="Constant_184465" type="Const" version="opset1">
335
  <data element_type="i32" shape="432" offset="7601028" size="1728" />
336
  <output>
337
  <port id="0" precision="I32">
 
339
  </port>
340
  </output>
341
  </layer>
342
+ <layer id="31" name="BPETokenizer_184466" type="BPETokenizer" version="extension">
343
  <data unk_token="" fuse_unk="false" suffix_indicator="" end_suffix="" byte_fallback="false" cache_capacity="40003" />
344
  <input>
345
  <port id="0" precision="I32">
 
409
  </port>
410
  </output>
411
  </layer>
412
+ <layer id="32" name="Constant_184467" type="Const" version="opset1">
413
  <data element_type="i32" shape="" offset="7602756" size="4" />
414
  <output>
415
  <port id="0" precision="I32" />
416
  </output>
417
  </layer>
418
+ <layer id="33" name="Constant_184469" type="Const" version="opset1">
419
  <data element_type="u8" shape="4" offset="7602760" size="4" />
420
  <output>
421
  <port id="0" precision="U8">
 
423
  </port>
424
  </output>
425
  </layer>
426
+ <layer id="34" name="Constant_184471" type="Const" version="opset1">
427
  <data element_type="u8" shape="13" offset="7602764" size="13" />
428
  <output>
429
  <port id="0" precision="U8">
 
431
  </port>
432
  </output>
433
  </layer>
434
+ <layer id="35" name="Truncate_184472" type="Truncate" version="extension">
435
  <data m_num_inputs="1" />
436
  <input>
437
  <port id="0" precision="I32">
 
463
  </port>
464
  </output>
465
  </layer>
466
+ <layer id="36" name="Subtract_184473" type="Subtract" version="opset1">
467
  <data auto_broadcast="numpy" />
468
  <input>
469
  <port id="0" precision="I32">
 
479
  </port>
480
  </output>
481
  </layer>
482
+ <layer id="37" name="Constant_184474" type="Const" version="opset1">
483
  <data element_type="i32" shape="" offset="7602777" size="4" />
484
  <output>
485
  <port id="0" precision="I32" />
486
  </output>
487
  </layer>
488
+ <layer id="38" name="ReduceMax_184475" type="ReduceMax" version="opset1">
489
  <data keep_dims="false" />
490
  <input>
491
  <port id="0" precision="I32">
 
497
  <port id="2" precision="I32" />
498
  </output>
499
  </layer>
500
+ <layer id="39" name="Constant_184476" type="Const" version="opset1">
501
  <data element_type="i32" shape="" offset="7602781" size="4" />
502
  <output>
503
  <port id="0" precision="I32" />
504
  </output>
505
  </layer>
506
+ <layer id="40" name="RaggedToDense_184477" type="RaggedToDense" version="extension">
507
  <data pad_right="false" m_pad_max_length="false" />
508
  <input>
509
  <port id="0" precision="I32">
 
529
  </port>
530
  </output>
531
  </layer>
532
+ <layer id="41" name="Convert_184478" type="Convert" version="opset1">
533
  <data destination_type="i32" />
534
  <input>
535
  <port id="0" precision="BOOL">
 
544
  </port>
545
  </output>
546
  </layer>
547
+ <layer id="42" name="Convert_184478.0" type="Convert" version="opset1">
548
  <data destination_type="i64" />
549
  <input>
550
  <port id="0" precision="I32">
 
559
  </port>
560
  </output>
561
  </layer>
562
+ <layer id="44" name="RaggedToDense_184477.0" type="Convert" version="opset1">
563
  <data destination_type="i64" />
564
  <input>
565
  <port id="0" precision="I32">
 
574
  </port>
575
  </output>
576
  </layer>
577
+ <layer id="45" name="Result_184479" type="Result" version="opset1" output_names="input_ids">
578
  <input>
579
  <port id="0" precision="I64">
580
  <dim>-1</dim>
 
582
  </port>
583
  </input>
584
  </layer>
585
+ <layer id="43" name="Result_184480" type="Result" version="opset1" output_names="attention_mask">
586
  <input>
587
  <port id="0" precision="I64">
588
  <dim>-1</dim>
 
669
  <handle_special_tokens_with_re />
670
  <max_length />
671
  <number_of_inputs value="1" />
672
+ <openvino_tokenizers_version value="2025.2.0.0-566-85be884a69f" />
673
+ <openvino_version value="2025.2.0-19140-c01cd93e24d-releases/2025/2" />
674
  <original_tokenizer_class value="&lt;class 'transformers.models.gpt2.tokenization_gpt2_fast.GPT2TokenizerFast'>" />
675
  <pad_token_id value="199999" />
 
676
  <skip_special_tokens value="True" />
677
  <streaming_detokenizer value="False" />
 
678
  <tokenizer_output_type value="i64" />
679
  <tokenizers_version value="0.21.1" />
680
  <transformers_version value="4.51.3" />