Hugging Face's logo

fakezeta
/
Phi-3-medium-4k-instruct-ov-int4

Text Generation
Transformers
OpenVINO
multilingual
phi3
nlp
code
conversational
custom_code
Inference Endpoints
Model card Files Community
fakezeta commited on
Commit
8893362
•
1 Parent(s): 4c1a282

Upload 13 files

Browse files

Converted with transformers 4.41.2

Files changed (5) hide show
  1. config.json +1 -1
  2. generation_config.json +1 -1
  3. openvino_detokenizer.xml +7 -7
  4. openvino_model.xml +0 -0
  5. openvino_tokenizer.xml +15 -15
config.json CHANGED
@@ -29,7 +29,7 @@
29
  "rope_theta": 10000.0,
30
  "sliding_window": 2047,
31
  "tie_word_embeddings": false,
32
- "transformers_version": "4.40.2",
33
  "use_cache": true,
34
  "vocab_size": 32064
35
  }
 
29
  "rope_theta": 10000.0,
30
  "sliding_window": 2047,
31
  "tie_word_embeddings": false,
32
+ "transformers_version": "4.41.2",
33
  "use_cache": true,
34
  "vocab_size": 32064
35
  }
generation_config.json CHANGED
@@ -7,5 +7,5 @@
7
  32007
8
  ],
9
  "pad_token_id": 32000,
10
- "transformers_version": "4.40.2"
11
  }
 
7
  32007
8
  ],
9
  "pad_token_id": 32000,
10
+ "transformers_version": "4.41.2"
11
  }
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_305146" type="Parameter" version="opset1">
5
  <data shape="?,?" element_type="i64" />
6
  <output>
7
- <port id="0" precision="I64" names="Parameter_305146">
8
  <dim>-1</dim>
9
  <dim>-1</dim>
10
  </port>
11
  </output>
12
  </layer>
13
- <layer id="1" name="Constant_305126" type="Const" version="opset1">
14
  <data element_type="u8" shape="499969" offset="0" size="499969" />
15
  <output>
16
  <port id="0" precision="U8">
@@ -18,7 +18,7 @@
18
  </port>
19
  </output>
20
  </layer>
21
- <layer id="2" name="Convert_305156" 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_305147" type="SentencepieceDetokenizer" version="extension">
37
  <input>
38
  <port id="0" precision="U8">
39
  <dim>499969</dim>
@@ -55,7 +55,7 @@
55
  </port>
56
  </output>
57
  </layer>
58
- <layer id="4" name="StringTensorPack_305148" 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_305149" type="Result" version="opset1">
78
  <input>
79
  <port id="0" precision="STRING">
80
  <dim>-1</dim>
 
1
  <?xml version="1.0"?>
2
  <net name="detokenizer" version="11">
3
  <layers>
4
+ <layer id="0" name="Parameter_305226" type="Parameter" version="opset1">
5
  <data shape="?,?" element_type="i64" />
6
  <output>
7
+ <port id="0" precision="I64" names="Parameter_305226">
8
  <dim>-1</dim>
9
  <dim>-1</dim>
10
  </port>
11
  </output>
12
  </layer>
13
+ <layer id="1" name="Constant_305206" type="Const" version="opset1">
14
  <data element_type="u8" shape="499969" offset="0" size="499969" />
15
  <output>
16
  <port id="0" precision="U8">
 
18
  </port>
19
  </output>
20
  </layer>
21
+ <layer id="2" name="Convert_305236" 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_305227" type="SentencepieceDetokenizer" version="extension">
37
  <input>
38
  <port id="0" precision="U8">
39
  <dim>499969</dim>
 
55
  </port>
56
  </output>
57
  </layer>
58
+ <layer id="4" name="StringTensorPack_305228" 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_305229" type="Result" version="opset1">
78
  <input>
79
  <port id="0" precision="STRING">
80
  <dim>-1</dim>
openvino_model.xml CHANGED
The diff for this file is too large to render. See raw diff
 
openvino_tokenizer.xml CHANGED
@@ -9,13 +9,13 @@
9
  </port>
10
  </output>
11
  </layer>
12
- <layer id="1" name="Constant_305132" type="Const" version="opset1">
13
  <data element_type="i32" shape="" offset="0" size="4" />
14
  <output>
15
  <port id="0" precision="I32" />
16
  </output>
17
  </layer>
18
- <layer id="2" name="Constant_305126" type="Const" version="opset1">
19
  <data element_type="u8" shape="499969" offset="4" size="499969" />
20
  <output>
21
  <port id="0" precision="U8">
@@ -23,7 +23,7 @@
23
  </port>
24
  </output>
25
  </layer>
26
- <layer id="3" name="SentencepieceTokenizer_305128" type="SentencepieceTokenizer" version="extension">
27
  <data nbest_size="0" alpha="0" add_bos="false" add_eos="false" reverse="false" />
28
  <input>
29
  <port id="0" precision="U8">
@@ -46,7 +46,7 @@
46
  </port>
47
  </output>
48
  </layer>
49
- <layer id="4" name="Broadcast_305133" type="Broadcast" version="opset3">
50
  <data mode="numpy" />
51
  <input>
52
  <port id="0" precision="I32" />
@@ -61,13 +61,13 @@
61
  </port>
62
  </output>
63
  </layer>
64
- <layer id="5" name="Constant_305134" 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_305135" type="ShapeOf" version="opset3">
71
  <data output_type="i64" />
72
  <input>
73
  <port id="0" precision="I32">
@@ -80,7 +80,7 @@
80
  </port>
81
  </output>
82
  </layer>
83
- <layer id="7" name="Broadcast_305136" type="Broadcast" version="opset3">
84
  <data mode="numpy" />
85
  <input>
86
  <port id="0" precision="I32" />
@@ -94,7 +94,7 @@
94
  </port>
95
  </output>
96
  </layer>
97
- <layer id="8" name="ScatterNDUpdate_305140" type="ScatterNDUpdate" version="opset4">
98
  <input>
99
  <port id="0" precision="I32">
100
  <dim>-1</dim>
@@ -115,7 +115,7 @@
115
  </port>
116
  </output>
117
  </layer>
118
- <layer id="9" name="ScatterNDUpdate_305140" type="Convert" version="opset1">
119
  <data destination_type="i64" />
120
  <input>
121
  <port id="0" precision="I32">
@@ -130,13 +130,13 @@
130
  </port>
131
  </output>
132
  </layer>
133
- <layer id="11" name="Constant_305129" 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_305130" type="Broadcast" version="opset3">
140
  <data mode="numpy" />
141
  <input>
142
  <port id="0" precision="I32" />
@@ -151,7 +151,7 @@
151
  </port>
152
  </output>
153
  </layer>
154
- <layer id="13" name="ScatterNDUpdate_305131" type="ScatterNDUpdate" version="opset4">
155
  <input>
156
  <port id="0" precision="I32">
157
  <dim>-1</dim>
@@ -172,7 +172,7 @@
172
  </port>
173
  </output>
174
  </layer>
175
- <layer id="14" name="ScatterNDUpdate_305131" type="Convert" version="opset1">
176
  <data destination_type="i64" />
177
  <input>
178
  <port id="0" precision="I32">
@@ -187,7 +187,7 @@
187
  </port>
188
  </output>
189
  </layer>
190
- <layer id="15" name="Result_305141" type="Result" version="opset1">
191
  <input>
192
  <port id="0" precision="I64">
193
  <dim>-1</dim>
@@ -195,7 +195,7 @@
195
  </port>
196
  </input>
197
  </layer>
198
- <layer id="10" name="Result_305142" type="Result" version="opset1">
199
  <input>
200
  <port id="0" precision="I64">
201
  <dim>-1</dim>
 
9
  </port>
10
  </output>
11
  </layer>
12
+ <layer id="1" name="Constant_305212" type="Const" version="opset1">
13
  <data element_type="i32" shape="" offset="0" size="4" />
14
  <output>
15
  <port id="0" precision="I32" />
16
  </output>
17
  </layer>
18
+ <layer id="2" name="Constant_305206" type="Const" version="opset1">
19
  <data element_type="u8" shape="499969" offset="4" size="499969" />
20
  <output>
21
  <port id="0" precision="U8">
 
23
  </port>
24
  </output>
25
  </layer>
26
+ <layer id="3" name="SentencepieceTokenizer_305208" type="SentencepieceTokenizer" version="extension">
27
  <data nbest_size="0" alpha="0" add_bos="false" add_eos="false" reverse="false" />
28
  <input>
29
  <port id="0" precision="U8">
 
46
  </port>
47
  </output>
48
  </layer>
49
+ <layer id="4" name="Broadcast_305213" type="Broadcast" version="opset3">
50
  <data mode="numpy" />
51
  <input>
52
  <port id="0" precision="I32" />
 
61
  </port>
62
  </output>
63
  </layer>
64
+ <layer id="5" name="Constant_305214" 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_305215" type="ShapeOf" version="opset3">
71
  <data output_type="i64" />
72
  <input>
73
  <port id="0" precision="I32">
 
80
  </port>
81
  </output>
82
  </layer>
83
+ <layer id="7" name="Broadcast_305216" type="Broadcast" version="opset3">
84
  <data mode="numpy" />
85
  <input>
86
  <port id="0" precision="I32" />
 
94
  </port>
95
  </output>
96
  </layer>
97
+ <layer id="8" name="ScatterNDUpdate_305220" type="ScatterNDUpdate" version="opset4">
98
  <input>
99
  <port id="0" precision="I32">
100
  <dim>-1</dim>
 
115
  </port>
116
  </output>
117
  </layer>
118
+ <layer id="9" name="ScatterNDUpdate_305220" type="Convert" version="opset1">
119
  <data destination_type="i64" />
120
  <input>
121
  <port id="0" precision="I32">
 
130
  </port>
131
  </output>
132
  </layer>
133
+ <layer id="11" name="Constant_305209" 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_305210" type="Broadcast" version="opset3">
140
  <data mode="numpy" />
141
  <input>
142
  <port id="0" precision="I32" />
 
151
  </port>
152
  </output>
153
  </layer>
154
+ <layer id="13" name="ScatterNDUpdate_305211" type="ScatterNDUpdate" version="opset4">
155
  <input>
156
  <port id="0" precision="I32">
157
  <dim>-1</dim>
 
172
  </port>
173
  </output>
174
  </layer>
175
+ <layer id="14" name="ScatterNDUpdate_305211" type="Convert" version="opset1">
176
  <data destination_type="i64" />
177
  <input>
178
  <port id="0" precision="I32">
 
187
  </port>
188
  </output>
189
  </layer>
190
+ <layer id="15" name="Result_305221" type="Result" version="opset1">
191
  <input>
192
  <port id="0" precision="I64">
193
  <dim>-1</dim>
 
195
  </port>
196
  </input>
197
  </layer>
198
+ <layer id="10" name="Result_305222" type="Result" version="opset1">
199
  <input>
200
  <port id="0" precision="I64">
201
  <dim>-1</dim>