Hugging Face's logo

SawanStack
/
gpt2-image-captioning-onnx

Image-Text-to-Text
Transformers
ONNX
vision-encoder-decoder
Model card Files Community
SawanStack commited on
Commit
ceb48a1
·
1 Parent(s): 08729f4

Upload 16 files

Browse files
Files changed (12) hide show
  1. config.json +1 -1
  2. generation_config.json +1 -1
  3. onnx/decoder_model.onnx +3 -0
  4. onnx/decoder_model_merged.onnx +3 -0
  5. onnx/decoder_model_merged_quantized.onnx +3 -0
  6. onnx/decoder_model_quantized.onnx +3 -0
  7. onnx/decoder_with_past_model.onnx +3 -0
  8. onnx/decoder_with_past_model_quantized.onnx +3 -0
  9. onnx/encoder_model.onnx +3 -0
  10. onnx/encoder_model_quantized.onnx +3 -0
  11. quantize_config.json +125 -0
  12. tokenizer_config.json +0 -10
config.json CHANGED
@@ -181,5 +181,5 @@
181
  "model_type": "vision-encoder-decoder",
182
  "pad_token_id": 50256,
183
  "tie_word_embeddings": false,
184
- "transformers_version": "4.35.0"
185
  }
 
181
  "model_type": "vision-encoder-decoder",
182
  "pad_token_id": 50256,
183
  "tie_word_embeddings": false,
184
+ "transformers_version": "4.33.2"
185
  }
generation_config.json CHANGED
@@ -4,5 +4,5 @@
4
  "decoder_start_token_id": 50256,
5
  "eos_token_id": 50256,
6
  "pad_token_id": 50256,
7
- "transformers_version": "4.35.0"
8
  }
 
4
  "decoder_start_token_id": 50256,
5
  "eos_token_id": 50256,
6
  "pad_token_id": 50256,
7
+ "transformers_version": "4.33.2"
8
  }
onnx/decoder_model.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ee134eb4fae915ae6b8fa9b7ece5a8413fd9327fc67631e8dd1b6ad462f611e
3
+ size 613129811
onnx/decoder_model_merged.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93414b5764f4827c724b1322f161ba7f1f3bceb7ebf8a7a640dcb204becc4e42
3
+ size 615025088
onnx/decoder_model_merged_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a77f15821d7e404019baf7ef18846fea48dd0aa39038671e0012eb5c7843e5b
3
+ size 158000536
onnx/decoder_model_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc4a92027f301ea63859420de31b1d815646d15dd1af309b1a099c87b1c9d92d
3
+ size 155678727
onnx/decoder_with_past_model.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba4c6f85d6ef778d5fceb5217191a660747e912471182ce772a95ae7ad179df2
3
+ size 613127119
onnx/decoder_with_past_model_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c97917f85b46734eb6da5a1ded2e9ac30161f0168a54fe761c7fedb60e55d3df
3
+ size 155670590
onnx/encoder_model.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7724bf28e2fdd680119c2ed0fdcdd612b0ce5543b77f20f2ff046fe797e5f756
3
+ size 343440632
onnx/encoder_model_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e34da798dfce2b4daaa5b1bf595fa7a67046a1848054565a4ff36056e227d801
3
+ size 87038171
quantize_config.json ADDED
@@ -0,0 +1,125 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": false,
3
+ "reduce_range": false,
4
+ "per_model_config": {
5
+ "decoder_with_past_model": {
6
+ "op_types": [
7
+ "Squeeze",
8
+ "Gemm",
9
+ "Softmax",
10
+ "Unsqueeze",
11
+ "ConstantOfShape",
12
+ "Mul",
13
+ "Slice",
14
+ "Div",
15
+ "Where",
16
+ "Tanh",
17
+ "Sub",
18
+ "Cast",
19
+ "Constant",
20
+ "Add",
21
+ "Transpose",
22
+ "MatMul",
23
+ "Shape",
24
+ "Pow",
25
+ "Concat",
26
+ "Sqrt",
27
+ "Range",
28
+ "Gather",
29
+ "Split",
30
+ "ReduceMean",
31
+ "Reshape"
32
+ ],
33
+ "weight_type": "QInt8"
34
+ },
35
+ "decoder_model": {
36
+ "op_types": [
37
+ "Squeeze",
38
+ "Gemm",
39
+ "Softmax",
40
+ "Unsqueeze",
41
+ "ConstantOfShape",
42
+ "Mul",
43
+ "Slice",
44
+ "Div",
45
+ "Where",
46
+ "Tanh",
47
+ "Sub",
48
+ "Cast",
49
+ "Constant",
50
+ "Add",
51
+ "Transpose",
52
+ "MatMul",
53
+ "Shape",
54
+ "Pow",
55
+ "Concat",
56
+ "Sqrt",
57
+ "Range",
58
+ "Gather",
59
+ "Split",
60
+ "ReduceMean",
61
+ "Reshape"
62
+ ],
63
+ "weight_type": "QInt8"
64
+ },
65
+ "encoder_model": {
66
+ "op_types": [
67
+ "Erf",
68
+ "Softmax",
69
+ "Unsqueeze",
70
+ "ConstantOfShape",
71
+ "Mul",
72
+ "Slice",
73
+ "Expand",
74
+ "Div",
75
+ "Where",
76
+ "Sub",
77
+ "Constant",
78
+ "Transpose",
79
+ "Add",
80
+ "MatMul",
81
+ "Shape",
82
+ "Equal",
83
+ "Pow",
84
+ "Concat",
85
+ "Sqrt",
86
+ "Conv",
87
+ "Gather",
88
+ "ReduceMean",
89
+ "Reshape"
90
+ ],
91
+ "weight_type": "QUInt8"
92
+ },
93
+ "decoder_model_merged": {
94
+ "op_types": [
95
+ "Squeeze",
96
+ "Gemm",
97
+ "Softmax",
98
+ "If",
99
+ "Unsqueeze",
100
+ "ConstantOfShape",
101
+ "Mul",
102
+ "Slice",
103
+ "Div",
104
+ "Where",
105
+ "Tanh",
106
+ "Sub",
107
+ "Cast",
108
+ "Constant",
109
+ "Add",
110
+ "Transpose",
111
+ "MatMul",
112
+ "Shape",
113
+ "Pow",
114
+ "Concat",
115
+ "Sqrt",
116
+ "Range",
117
+ "Gather",
118
+ "Split",
119
+ "ReduceMean",
120
+ "Reshape"
121
+ ],
122
+ "weight_type": "QInt8"
123
+ }
124
+ }
125
+ }
tokenizer_config.json CHANGED
@@ -1,15 +1,5 @@
1
  {
2
  "add_prefix_space": false,
3
- "added_tokens_decoder": {
4
- "50256": {
5
- "content": "<|endoftext|>",
6
- "lstrip": false,
7
- "normalized": false,
8
- "rstrip": false,
9
- "single_word": false,
10
- "special": true
11
- }
12
- },
13
  "bos_token": "<|endoftext|>",
14
  "clean_up_tokenization_spaces": true,
15
  "eos_token": "<|endoftext|>",
 
1
  {
2
  "add_prefix_space": false,
 
 
 
 
 
 
 
 
 
 
3
  "bos_token": "<|endoftext|>",
4
  "clean_up_tokenization_spaces": true,
5
  "eos_token": "<|endoftext|>",