Xenova HF staff commited on
Commit
3c39859
1 Parent(s): e7beaa9

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": 67028
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": 67028
60
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  67027
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 6,
14
  "pad_token_id": 67027,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  67027
 
11
  "max_length": 512,
12
  "num_beams": 6,
13
  "pad_token_id": 67027,
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 sha256:01286a6ed817b50f5c4e56b4a2f8856ee61d2bfb793926f6de712b37dd595001
3
- size 377036360
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:afb53d15902ff6e877b328d09703f7772cb2b320e26bb82a0cc209628e19f9de
3
+ size 239763073
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a36561d53736d762ea22c6037472dfa25a0d1043ec5298adde04dabc6c95de4c
3
- size 377271842
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92d017a39a5e24541fd93744c92c5c31c0e27591af8691cf327572e74586b7f1
3
+ size 239998640
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b5b9ea9e4fa20a2acc6606f46d98eb2da2b2fe200e69247d0c0c36d98efbbc22
3
- size 95913624
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad4df9a08f16952c0d9ee06807f54d62fcd07a9109c2c0b8ea05a4422d7bf4d7
3
+ size 61258736
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6cda797a25a777e5c997f50a78504bc73bec56183de9dc69f3a00179c56c715c
3
- size 95542268
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d70da1130c5a0746f8f18081854511b5a0671e563a56b4dd7870b2eb922944cc
3
+ size 60888034
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6644cf82e92b9344e80401290a8835ad0fe8c50f5c02b45f0bea462f4614095c
3
- size 364385851
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9740762a39621d0676c71ddbee0a4c61a26137f2a3ef4b95e080a853e60e8760
3
+ size 227112564
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3030cd58049b8b111bd7a929d9b4d58ed9e3547822f46220c18af62aefba0125
3
- size 92270804
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45526ea9df7953b44bf1c66fbe97d9da75ec0eb9eb3c3153af6a76a5ed16fcbf
3
+ size 57616570
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
  "Pow",
8
- "Expand",
9
- "Sub",
10
- "Sigmoid",
 
11
  "Softmax",
12
- "Constant",
13
- "ReduceMean",
 
14
  "Transpose",
15
- "ConstantOfShape",
16
- "Unsqueeze",
17
  "Reshape",
18
- "Range",
19
  "Concat",
20
- "MatMul",
21
- "Cast",
22
  "Equal",
23
- "Sqrt",
24
- "Div",
25
- "Gather",
26
- "Mul",
27
  "Add",
28
- "Where",
29
- "Shape"
 
 
 
 
 
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
- "decoder_with_past_model": {
34
  "op_types": [
 
35
  "Pow",
36
- "Expand",
37
- "Sub",
38
- "Sigmoid",
39
  "Softmax",
40
- "Constant",
41
- "ReduceMean",
42
  "Transpose",
43
- "ConstantOfShape",
44
- "Unsqueeze",
45
  "Reshape",
46
- "Range",
47
  "Concat",
48
- "MatMul",
49
- "Cast",
50
  "Equal",
51
- "Sqrt",
52
- "Div",
53
- "Gather",
54
- "Mul",
55
  "Add",
56
- "Where",
57
- "Shape"
 
 
 
 
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
 
63
  "Pow",
64
- "Expand",
65
- "Sub",
66
- "Sigmoid",
67
  "Squeeze",
 
 
 
 
68
  "Softmax",
69
- "Slice",
70
- "Constant",
71
- "ReduceMean",
72
  "Transpose",
73
- "ConstantOfShape",
74
- "Unsqueeze",
75
  "Reshape",
76
- "Less",
77
- "Range",
78
  "Concat",
79
- "MatMul",
80
- "Cast",
81
  "Equal",
82
- "Sqrt",
83
- "Div",
84
- "Gather",
85
- "Mul",
86
  "Add",
87
- "Where",
88
- "Shape"
 
 
 
 
 
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
 
94
  "Pow",
95
- "Expand",
96
- "Sub",
97
- "Sigmoid",
98
- "Squeeze",
99
  "Softmax",
100
- "Slice",
101
- "Constant",
102
- "ReduceMean",
103
  "Transpose",
104
- "ConstantOfShape",
105
- "Unsqueeze",
106
  "Reshape",
107
- "Less",
108
- "Range",
109
  "Concat",
110
  "Equal",
111
- "Cast",
112
- "MatMul",
113
- "Sqrt",
114
- "Div",
115
- "Gather",
116
- "Mul",
117
- "If",
118
  "Add",
119
- "Where",
120
- "Shape"
 
 
 
 
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
+ "Where",
8
  "Pow",
9
+ "Mul",
10
+ "Sqrt",
11
+ "Squeeze",
12
+ "Div",
13
  "Softmax",
14
+ "Cast",
15
+ "Gather",
16
+ "Less",
17
  "Transpose",
18
+ "MatMul",
 
19
  "Reshape",
 
20
  "Concat",
 
 
21
  "Equal",
22
+ "Expand",
23
+ "Unsqueeze",
24
+ "ConstantOfShape",
 
25
  "Add",
26
+ "Shape",
27
+ "ReduceMean",
28
+ "Sub",
29
+ "Slice",
30
+ "Range",
31
+ "Sigmoid",
32
+ "Constant"
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "encoder_model": {
37
  "op_types": [
38
+ "Where",
39
  "Pow",
40
+ "Mul",
41
+ "Sqrt",
42
+ "Div",
43
  "Softmax",
44
+ "Cast",
45
+ "Gather",
46
  "Transpose",
47
+ "MatMul",
 
48
  "Reshape",
 
49
  "Concat",
 
 
50
  "Equal",
51
+ "Expand",
52
+ "Unsqueeze",
53
+ "ConstantOfShape",
 
54
  "Add",
55
+ "Shape",
56
+ "ReduceMean",
57
+ "Sub",
58
+ "Range",
59
+ "Sigmoid",
60
+ "Constant"
61
  ],
62
  "weight_type": "QInt8"
63
  },
64
+ "decoder_model_merged": {
65
  "op_types": [
66
+ "Where",
67
  "Pow",
68
+ "Mul",
69
+ "Sqrt",
 
70
  "Squeeze",
71
+ "Div",
72
+ "If",
73
+ "Cast",
74
+ "Gather",
75
  "Softmax",
76
+ "Less",
 
 
77
  "Transpose",
78
+ "MatMul",
 
79
  "Reshape",
 
 
80
  "Concat",
 
 
81
  "Equal",
82
+ "Expand",
83
+ "Unsqueeze",
84
+ "ConstantOfShape",
 
85
  "Add",
86
+ "Shape",
87
+ "ReduceMean",
88
+ "Sub",
89
+ "Slice",
90
+ "Range",
91
+ "Sigmoid",
92
+ "Constant"
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
+ "decoder_with_past_model": {
97
  "op_types": [
98
+ "Where",
99
  "Pow",
100
+ "Mul",
101
+ "Sqrt",
102
+ "Div",
 
103
  "Softmax",
104
+ "Cast",
105
+ "Gather",
 
106
  "Transpose",
107
+ "MatMul",
 
108
  "Reshape",
 
 
109
  "Concat",
110
  "Equal",
111
+ "Expand",
112
+ "Unsqueeze",
113
+ "ConstantOfShape",
 
 
 
 
114
  "Add",
115
+ "Shape",
116
+ "ReduceMean",
117
+ "Sub",
118
+ "Range",
119
+ "Sigmoid",
120
+ "Constant"
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff