Xenova HF staff commited on
Commit
4c3e821
·
1 Parent(s): c006d4c

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -54,7 +54,7 @@
54
  "scale_embedding": true,
55
  "share_encoder_decoder_embeddings": true,
56
  "static_position_embeddings": true,
57
- "transformers_version": "4.32.0.dev0",
58
  "use_cache": true,
59
  "vocab_size": 62518
60
  }
 
54
  "scale_embedding": true,
55
  "share_encoder_decoder_embeddings": true,
56
  "static_position_embeddings": true,
57
+ "transformers_version": "4.34.0.dev0",
58
  "use_cache": true,
59
  "vocab_size": 62518
60
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  62517
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 6,
14
  "pad_token_id": 62517,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  62517
 
11
  "max_length": 512,
12
  "num_beams": 6,
13
  "pad_token_id": 62517,
14
+ "renormalize_logits": true,
15
+ "transformers_version": "4.34.0.dev0"
16
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:25e2daca91185fe64df60b906ed71bf1502c5ca29e5e70bf1f19dc750066f034
3
- size 358545360
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e396d82742dadb4ef1b8b2fdd7ec14487323d11027def76c765b5778218f7124
3
+ size 230508553
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:072efcc952c404ab36ed1c0a263fed685573c3b8152cf4564ffdb608098beaf1
3
- size 358780842
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6488b9b9591c5b61b221b14ab1fa5c0683da6a8c42071241ca36b63d82b5ab11
3
+ size 230744120
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b93089bee18c64446b7598da94e6020cba9035a8bcf9d11e139c92f5c86a1a10
3
- size 91254794
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25e7ff97b201bce64ab9c39749761d657b70b8602bcbc1ce01c67f713e90ffdf
3
+ size 58931576
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ac7d18a85c33d369d4620264daae4ff009db5cd145f0dbaefb7a4f3fd90e062f
3
- size 90883438
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d880abf58612013815554ef55c98d0f28b68190b2cb6f65c39d6e834f51dda8c
3
+ size 58560874
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7662f473d602635384d9c812d1c99c1375177ffe5db492b8d9c63b6eb2c1974d
3
- size 345894851
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6f0c5a4145df383395f4ee79b74529e2d7d8035139f25674ec9979eb7f8c730
3
+ size 217858044
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4382e6649711360ee7a041aa2171fb44c820c458b9fa7090e1181b5f230cc2b6
3
- size 87611974
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1596873df562c7b5d0b0a4da9021df3b8f06b23400d207a6d5071c6a624d3b2e
3
+ size 55289410
quantize_config.json CHANGED
@@ -2,122 +2,122 @@
2
  "per_channel": true,
3
  "reduce_range": true,
4
  "per_model_config": {
5
- "encoder_model": {
6
  "op_types": [
7
- "Cast",
8
- "Softmax",
9
- "Transpose",
10
- "Sigmoid",
11
- "Sub",
12
- "Unsqueeze",
13
- "Div",
14
- "Concat",
15
  "Range",
 
16
  "Where",
17
- "Shape",
18
- "MatMul",
19
- "Mul",
20
  "ConstantOfShape",
21
- "Reshape",
22
- "Pow",
23
- "Equal",
24
  "Sqrt",
 
 
 
25
  "Gather",
 
 
 
 
 
 
 
 
 
26
  "Add",
27
- "Constant",
28
- "Expand",
29
- "ReduceMean"
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
- "decoder_with_past_model": {
34
  "op_types": [
35
- "Cast",
36
- "Softmax",
37
- "Transpose",
38
- "Sigmoid",
39
- "Sub",
40
- "Unsqueeze",
41
- "Div",
42
- "Concat",
43
  "Range",
 
44
  "Where",
45
- "Shape",
46
- "MatMul",
47
- "Mul",
48
  "ConstantOfShape",
49
- "Reshape",
50
- "Pow",
51
- "Equal",
52
  "Sqrt",
 
 
53
  "Gather",
 
 
 
 
 
 
 
 
54
  "Add",
55
- "Constant",
56
- "Expand",
57
- "ReduceMean"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
63
- "Cast",
64
- "Softmax",
65
  "Squeeze",
66
- "Transpose",
67
- "Sigmoid",
68
- "Sub",
69
- "Unsqueeze",
70
- "Div",
71
- "Concat",
72
  "Range",
 
 
73
  "Where",
74
- "Slice",
75
- "Shape",
76
- "MatMul",
77
- "Less",
78
- "Mul",
79
  "ConstantOfShape",
80
- "Reshape",
81
- "Pow",
82
- "Equal",
83
  "Sqrt",
 
 
 
84
  "Gather",
 
 
 
 
 
 
 
 
 
85
  "Add",
86
- "Constant",
87
- "Expand",
88
- "ReduceMean"
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
94
- "Cast",
95
- "Softmax",
96
- "Squeeze",
97
- "Transpose",
98
- "Sigmoid",
99
- "Sub",
100
- "Unsqueeze",
101
- "If",
102
- "Div",
103
- "Concat",
104
  "Range",
 
105
  "Where",
106
- "Slice",
107
- "Shape",
108
- "MatMul",
109
- "Less",
110
- "Mul",
111
  "ConstantOfShape",
112
- "Reshape",
113
- "Pow",
114
- "Equal",
115
  "Sqrt",
 
 
116
  "Gather",
 
 
 
 
 
 
 
 
117
  "Add",
118
- "Constant",
119
- "Expand",
120
- "ReduceMean"
121
  ],
122
  "weight_type": "QInt8"
123
  }
 
2
  "per_channel": true,
3
  "reduce_range": true,
4
  "per_model_config": {
5
+ "decoder_model": {
6
  "op_types": [
7
+ "Squeeze",
8
+ "Expand",
9
+ "Reshape",
 
 
 
 
 
10
  "Range",
11
+ "Constant",
12
  "Where",
13
+ "Div",
14
+ "Cast",
 
15
  "ConstantOfShape",
 
 
 
16
  "Sqrt",
17
+ "Less",
18
+ "Unsqueeze",
19
+ "Softmax",
20
  "Gather",
21
+ "Shape",
22
+ "Equal",
23
+ "Concat",
24
+ "Sub",
25
+ "ReduceMean",
26
+ "Transpose",
27
+ "Pow",
28
+ "Sigmoid",
29
+ "Slice",
30
  "Add",
31
+ "MatMul",
32
+ "Mul"
 
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "encoder_model": {
37
  "op_types": [
38
+ "Expand",
39
+ "Reshape",
 
 
 
 
 
 
40
  "Range",
41
+ "Constant",
42
  "Where",
43
+ "Div",
44
+ "Cast",
 
45
  "ConstantOfShape",
 
 
 
46
  "Sqrt",
47
+ "Unsqueeze",
48
+ "Softmax",
49
  "Gather",
50
+ "Shape",
51
+ "Equal",
52
+ "Concat",
53
+ "Sub",
54
+ "ReduceMean",
55
+ "Transpose",
56
+ "Pow",
57
+ "Sigmoid",
58
  "Add",
59
+ "MatMul",
60
+ "Mul"
 
61
  ],
62
  "weight_type": "QInt8"
63
  },
64
+ "decoder_model_merged": {
65
  "op_types": [
 
 
66
  "Squeeze",
67
+ "Expand",
68
+ "Reshape",
 
 
 
 
69
  "Range",
70
+ "If",
71
+ "Constant",
72
  "Where",
73
+ "Div",
74
+ "Cast",
 
 
 
75
  "ConstantOfShape",
 
 
 
76
  "Sqrt",
77
+ "Less",
78
+ "Unsqueeze",
79
+ "Softmax",
80
  "Gather",
81
+ "Shape",
82
+ "Equal",
83
+ "Concat",
84
+ "Sub",
85
+ "ReduceMean",
86
+ "Transpose",
87
+ "Pow",
88
+ "Sigmoid",
89
+ "Slice",
90
  "Add",
91
+ "MatMul",
92
+ "Mul"
 
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
+ "decoder_with_past_model": {
97
  "op_types": [
98
+ "Expand",
99
+ "Reshape",
 
 
 
 
 
 
 
 
100
  "Range",
101
+ "Constant",
102
  "Where",
103
+ "Div",
104
+ "Cast",
 
 
 
105
  "ConstantOfShape",
 
 
 
106
  "Sqrt",
107
+ "Unsqueeze",
108
+ "Softmax",
109
  "Gather",
110
+ "Shape",
111
+ "Equal",
112
+ "Concat",
113
+ "Sub",
114
+ "ReduceMean",
115
+ "Transpose",
116
+ "Pow",
117
+ "Sigmoid",
118
  "Add",
119
+ "MatMul",
120
+ "Mul"
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff