Xenova HF staff commited on
Commit
5e7adf6
1 Parent(s): a1d8648

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -55,7 +55,7 @@
55
  "scale_embedding": true,
56
  "share_encoder_decoder_embeddings": true,
57
  "static_position_embeddings": true,
58
- "transformers_version": "4.32.0.dev0",
59
  "use_cache": true,
60
  "vocab_size": 58101
61
  }
 
55
  "scale_embedding": true,
56
  "share_encoder_decoder_embeddings": true,
57
  "static_position_embeddings": true,
58
+ "transformers_version": "4.34.0.dev0",
59
  "use_cache": true,
60
  "vocab_size": 58101
61
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  58100
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 4,
14
  "pad_token_id": 58100,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  58100
 
11
  "max_length": 512,
12
  "num_beams": 4,
13
  "pad_token_id": 58100,
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:b89e5396c0b6862b687a562cae33b431ce5567842531ac5c0c32d170ccb1e1a1
3
- size 340435660
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a642689a46e4b140cee262ab6e775f4c84427b7e3d567b8cbbc32989a771244d
3
+ size 221444869
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f441e29e613e354d548fa8b3a389f624b807bc40cebde161fdeef390868ed9db
3
- size 340671142
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bea9bca485118db5f5300bd1c77e8e07c9163215b9b5e1cf36d1e81b39cb57cb
3
+ size 221680436
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fbae2d8992d3fca2a22b4e3259e9dc746b21266f44fee68d051faff3d23da849
3
- size 86692033
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b46a825964cdd182fe47cc780f0fb87d357eae1f47cf20aa63c1c09be5b510c
3
+ size 56652404
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:80419ebbad2984c725ce20a84923110febe0307d35aff5c61874daa6b89b1ef5
3
- size 86320677
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75ef79aa9bde9e3dce9ca584c29507be5f464973f6c600c89ff419bc8de29ebc
3
+ size 56281702
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4bcf2bac08eee3fafe46990ff70b8d6ce99d19599c0c697d56ed9afc3febd3ea
3
- size 327785151
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2feed7ef7ca0ffc6d1ff1c1f102c7e932b1bfc8f51bcbc345da116502a71adf
3
+ size 208794360
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a665756e6531732ed79140ab2dac107197a175cdfb5d5787f1bca1a30e5289fc
3
- size 83049213
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e76293c9b29317d09e80dc2dd721092b7216367e12890d1baff5255cd9210c2a
3
+ size 53010238
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
  "Expand",
 
8
  "Where",
9
- "ConstantOfShape",
10
- "Shape",
11
- "Cast",
12
- "Equal",
13
  "Pow",
14
- "Range",
15
- "Softmax",
16
- "Unsqueeze",
17
  "Reshape",
 
 
18
  "MatMul",
 
19
  "Sqrt",
20
- "Add",
21
- "Constant",
22
  "ReduceMean",
 
 
 
 
23
  "Mul",
24
- "Gather",
25
- "Sub",
26
- "Concat",
27
- "Transpose",
28
- "Sigmoid",
29
- "Div"
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
- "decoder_with_past_model": {
34
  "op_types": [
 
 
35
  "Expand",
 
36
  "Where",
37
- "ConstantOfShape",
38
- "Shape",
39
- "Cast",
40
- "Equal",
41
  "Pow",
42
- "Range",
43
- "Softmax",
44
- "Unsqueeze",
45
  "Reshape",
 
 
46
  "MatMul",
47
  "Sqrt",
48
- "Add",
49
- "Constant",
50
  "ReduceMean",
 
 
 
 
51
  "Mul",
52
- "Gather",
53
- "Sub",
54
- "Concat",
55
- "Transpose",
56
- "Sigmoid",
57
- "Div"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
 
 
 
 
63
  "Expand",
 
64
  "Where",
65
- "ConstantOfShape",
66
- "Shape",
67
- "Cast",
68
- "Less",
69
- "Equal",
70
  "Pow",
71
- "Range",
72
- "Slice",
73
- "Softmax",
74
- "Unsqueeze",
75
- "Squeeze",
76
  "Reshape",
 
 
77
  "MatMul",
 
78
  "Sqrt",
79
- "Add",
80
- "Constant",
81
  "ReduceMean",
 
 
 
 
82
  "Mul",
83
- "Gather",
84
- "Sub",
85
- "Concat",
86
- "Transpose",
87
- "Sigmoid",
88
- "Div"
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
 
 
94
  "Expand",
 
95
  "Where",
96
- "ConstantOfShape",
97
- "Shape",
98
- "Cast",
99
- "Less",
100
- "Equal",
101
  "Pow",
102
- "Range",
103
- "If",
104
- "Slice",
105
- "Softmax",
106
- "Unsqueeze",
107
- "Squeeze",
108
  "Reshape",
 
 
109
  "MatMul",
110
  "Sqrt",
111
- "Add",
112
- "Constant",
113
  "ReduceMean",
 
 
 
 
114
  "Mul",
115
- "Gather",
116
- "Sub",
117
- "Concat",
118
- "Transpose",
119
- "Sigmoid",
120
- "Div"
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
+ "ConstantOfShape",
8
+ "Gather",
9
+ "Squeeze",
10
  "Expand",
11
+ "Transpose",
12
  "Where",
13
+ "Concat",
 
 
 
14
  "Pow",
15
+ "Less",
16
+ "Sub",
 
17
  "Reshape",
18
+ "Add",
19
+ "Div",
20
  "MatMul",
21
+ "Slice",
22
  "Sqrt",
23
+ "Sigmoid",
 
24
  "ReduceMean",
25
+ "Cast",
26
+ "Constant",
27
+ "Shape",
28
+ "Softmax",
29
  "Mul",
30
+ "Range",
31
+ "Equal",
32
+ "Unsqueeze"
 
 
 
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "encoder_model": {
37
  "op_types": [
38
+ "ConstantOfShape",
39
+ "Gather",
40
  "Expand",
41
+ "Transpose",
42
  "Where",
43
+ "Concat",
 
 
 
44
  "Pow",
45
+ "Sub",
 
 
46
  "Reshape",
47
+ "Add",
48
+ "Div",
49
  "MatMul",
50
  "Sqrt",
51
+ "Sigmoid",
 
52
  "ReduceMean",
53
+ "Cast",
54
+ "Constant",
55
+ "Shape",
56
+ "Softmax",
57
  "Mul",
58
+ "Range",
59
+ "Equal",
60
+ "Unsqueeze"
 
 
 
61
  ],
62
  "weight_type": "QInt8"
63
  },
64
+ "decoder_model_merged": {
65
  "op_types": [
66
+ "ConstantOfShape",
67
+ "If",
68
+ "Gather",
69
+ "Squeeze",
70
  "Expand",
71
+ "Transpose",
72
  "Where",
73
+ "Concat",
 
 
 
 
74
  "Pow",
75
+ "Less",
76
+ "Sub",
 
 
 
77
  "Reshape",
78
+ "Add",
79
+ "Div",
80
  "MatMul",
81
+ "Slice",
82
  "Sqrt",
83
+ "Sigmoid",
 
84
  "ReduceMean",
85
+ "Cast",
86
+ "Constant",
87
+ "Shape",
88
+ "Softmax",
89
  "Mul",
90
+ "Range",
91
+ "Equal",
92
+ "Unsqueeze"
 
 
 
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
+ "decoder_with_past_model": {
97
  "op_types": [
98
+ "ConstantOfShape",
99
+ "Gather",
100
  "Expand",
101
+ "Transpose",
102
  "Where",
103
+ "Concat",
 
 
 
 
104
  "Pow",
105
+ "Sub",
 
 
 
 
 
106
  "Reshape",
107
+ "Add",
108
+ "Div",
109
  "MatMul",
110
  "Sqrt",
111
+ "Sigmoid",
 
112
  "ReduceMean",
113
+ "Cast",
114
+ "Constant",
115
+ "Shape",
116
+ "Softmax",
117
  "Mul",
118
+ "Range",
119
+ "Equal",
120
+ "Unsqueeze"
 
 
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff