Xenova HF staff commited on
Commit
0ecf243
1 Parent(s): 9ef192e

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": 80379
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": 80379
60
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  80378
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 6,
14
  "pad_token_id": 80378,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  80378
 
11
  "max_length": 512,
12
  "num_beams": 6,
13
  "pad_token_id": 80378,
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:a359dc3dca5603297cf187fc32208b9ab2600a1e201c1e9ca985e326081e884a
3
- size 431775460
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef237c4ff5a2ecb197609233a84843a42bc900e283e8d1a8cb376ea2042465bc
3
+ size 267159325
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:062c587fe59cb2b8b7befccacd7d2248dcf3064f6db21e7228a02d50073352fc
3
- size 432010942
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62bebc21b22a969a674458bb26a2b2188e0fc4e89321df3ef0890d7902643152
3
+ size 267394892
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ab140f973e212b51dc7cc1521aad0ed92d6102d7ff21ca594b453638f77f4cec
3
- size 109705207
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aebc2789cd2ef8062fbda020bac90b61e68458ba98a4c43cea63c1d2310ba2b8
3
+ size 68147852
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3613b01af756c2086cdade24d074c303045dca80036b9689f7fb3c87752586e8
3
- size 109333851
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:203bac7b98e438e65e1ddfcbec254ffb6811e6a8f4f71e93517ca9550323bf6e
3
+ size 67777150
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e5421b8fa00730bb13c7d75ff90a4598306a6be8235e6dba9294488b344a7c3a
3
- size 419124951
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:928c45ed46779d7fd0eea2d47172eb79f2d9f44fa052be36236b7444735e9173
3
+ size 254508816
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6206ba77a84434c3f15b552ddfcd7fc4948f4d83aefb799190633666b58217dc
3
- size 106062387
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a209e9c09ce1c35e22f7cde57b42aa5b7373f2d9b549c0a9f836f2ff07748a6
3
+ size 64505686
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
- "Div",
8
- "ReduceMean",
9
- "Softmax",
10
- "Expand",
11
- "Unsqueeze",
12
- "Sqrt",
13
- "Sigmoid",
14
- "Constant",
15
- "Mul",
16
- "Where",
17
- "Range",
18
- "Equal",
19
  "Add",
 
 
 
 
 
20
  "Pow",
21
- "Cast",
22
  "ConstantOfShape",
23
  "Sub",
24
- "Reshape",
25
- "Shape",
26
- "Transpose",
 
 
27
  "MatMul",
28
- "Concat",
29
- "Gather"
 
 
 
 
 
 
 
 
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
- "decoder_with_past_model": {
34
  "op_types": [
35
- "Div",
36
- "ReduceMean",
37
- "Softmax",
38
- "Expand",
39
- "Unsqueeze",
40
- "Sqrt",
41
- "Sigmoid",
42
- "Constant",
43
- "Mul",
44
- "Where",
45
- "Range",
46
- "Equal",
47
  "Add",
 
 
 
48
  "Pow",
49
- "Cast",
 
50
  "ConstantOfShape",
51
  "Sub",
52
- "Reshape",
 
 
 
 
 
 
 
53
  "Shape",
 
 
54
  "Transpose",
55
- "MatMul",
56
- "Concat",
57
- "Gather"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
63
- "Div",
64
- "ReduceMean",
65
- "Softmax",
66
- "Expand",
67
- "Unsqueeze",
68
- "Sqrt",
69
- "Sigmoid",
70
- "Constant",
71
- "Mul",
72
- "Slice",
73
- "Where",
74
- "Range",
75
- "Equal",
76
  "Add",
 
 
 
77
  "Squeeze",
78
- "Pow",
79
  "Less",
80
- "Cast",
 
 
81
  "ConstantOfShape",
82
  "Sub",
83
- "Reshape",
84
- "Shape",
85
- "Transpose",
 
 
86
  "MatMul",
87
- "Concat",
88
- "Gather"
 
 
 
 
 
 
 
 
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
94
- "Div",
95
- "ReduceMean",
96
- "Softmax",
97
- "Expand",
98
- "Unsqueeze",
99
- "Sqrt",
100
- "Sigmoid",
101
- "Constant",
102
- "If",
103
- "Mul",
104
- "Slice",
105
- "Where",
106
- "Range",
107
  "Equal",
108
  "Add",
109
- "Squeeze",
 
110
  "Pow",
111
- "Less",
112
- "Cast",
113
  "ConstantOfShape",
114
  "Sub",
115
- "Reshape",
 
 
 
 
 
 
 
116
  "Shape",
 
 
117
  "Transpose",
118
- "MatMul",
119
- "Concat",
120
- "Gather"
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
  "Add",
8
+ "Equal",
9
+ "Mul",
10
+ "Constant",
11
+ "Squeeze",
12
+ "Less",
13
  "Pow",
14
+ "Concat",
15
  "ConstantOfShape",
16
  "Sub",
17
+ "Gather",
18
+ "Slice",
19
+ "Sqrt",
20
+ "Cast",
21
+ "Range",
22
  "MatMul",
23
+ "Expand",
24
+ "Softmax",
25
+ "Transpose",
26
+ "Sigmoid",
27
+ "Shape",
28
+ "Div",
29
+ "Unsqueeze",
30
+ "Reshape",
31
+ "ReduceMean",
32
+ "Where"
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "encoder_model": {
37
  "op_types": [
 
 
 
 
 
 
 
 
 
 
 
 
38
  "Add",
39
+ "Equal",
40
+ "Mul",
41
+ "Constant",
42
  "Pow",
43
+ "Concat",
44
+ "Reshape",
45
  "ConstantOfShape",
46
  "Sub",
47
+ "Gather",
48
+ "Sqrt",
49
+ "Cast",
50
+ "Range",
51
+ "MatMul",
52
+ "Expand",
53
+ "Softmax",
54
+ "Sigmoid",
55
  "Shape",
56
+ "Div",
57
+ "Unsqueeze",
58
  "Transpose",
59
+ "ReduceMean",
60
+ "Where"
 
61
  ],
62
  "weight_type": "QInt8"
63
  },
64
+ "decoder_model_merged": {
65
  "op_types": [
 
 
 
 
 
 
 
 
 
 
 
 
 
66
  "Add",
67
+ "Equal",
68
+ "Mul",
69
+ "Constant",
70
  "Squeeze",
 
71
  "Less",
72
+ "Pow",
73
+ "Concat",
74
+ "If",
75
  "ConstantOfShape",
76
  "Sub",
77
+ "Gather",
78
+ "Slice",
79
+ "Sqrt",
80
+ "Cast",
81
+ "Range",
82
  "MatMul",
83
+ "Expand",
84
+ "Softmax",
85
+ "Transpose",
86
+ "Sigmoid",
87
+ "Shape",
88
+ "Div",
89
+ "Unsqueeze",
90
+ "Reshape",
91
+ "ReduceMean",
92
+ "Where"
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
+ "decoder_with_past_model": {
97
  "op_types": [
 
 
 
 
 
 
 
 
 
 
 
 
 
98
  "Equal",
99
  "Add",
100
+ "Mul",
101
+ "Constant",
102
  "Pow",
103
+ "Concat",
104
+ "Reshape",
105
  "ConstantOfShape",
106
  "Sub",
107
+ "Gather",
108
+ "Sqrt",
109
+ "Cast",
110
+ "Range",
111
+ "MatMul",
112
+ "Expand",
113
+ "Softmax",
114
+ "Sigmoid",
115
  "Shape",
116
+ "Div",
117
+ "Unsqueeze",
118
  "Transpose",
119
+ "ReduceMean",
120
+ "Where"
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff