Xenova HF staff commited on
Commit
3274ee6
1 Parent(s): 90d3b82

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": 37745
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": 37745
60
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  37744
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 4,
14
  "pad_token_id": 37744,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  37744
 
11
  "max_length": 512,
12
  "num_beams": 4,
13
  "pad_token_id": 37744,
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:caad18fa59833b4ab7effcf5fd77ba9d8dec3c1d318ba5e98664b4ce47a16b82
3
- size 256976059
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99750cd9e5c6a6c6f492a03bf08532932fe2d8f00fe42297bb3dedf1935180e4
3
+ size 179674357
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f1f288e6ce9d4ca9f38305c90acc6c0ec1d85d7ba4aa5c8671f0889c26c4f749
3
- size 257211541
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:410f1cbcef22d8214b6bebf61f939d5bdf2f607e9ddb91ef9e1763af02e2a2b2
3
+ size 179909924
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b1dd6baa3bc101c080b8dc0fe607a0fce3b59de95c493cbce329820367ceba7f
3
- size 65664285
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c0291f2524b6c9843e32bc843f6a5f07db89ccb781caf15b325e5a25302e7ef
3
+ size 46148708
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c19b7aa331cee07571ae165a9cb4dcdc95b535afe63d3abccb76085bff653a24
3
- size 65292929
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe92f12d8b5733d0913d8ab1865d1dd0ca376c9ba9f4bd5ef37d725562e3288a
3
+ size 45778006
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:39a069ed8b04189564f3e27d1b46560703c3f6de976e239a5aaaf882e2a5d9a3
3
- size 244325550
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c681aab60969591cf1c4df987e650ab8bd0bab8d7e3a2c0adce1e4bb3f26009
3
+ size 167023848
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c1a572d7518233863a962db9a13104383d85556e9ef570f8593c5ee4dff28e90
3
- size 62021465
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38aee00bba12e88fd8a3712649ad6e9117dba4b804b3f6263abbab8c2340ca80
3
+ size 42506542
quantize_config.json CHANGED
@@ -4,120 +4,120 @@
4
  "per_model_config": {
5
  "encoder_model": {
6
  "op_types": [
7
- "Sqrt",
8
- "Shape",
9
  "Mul",
10
- "Pow",
11
- "Range",
12
  "Constant",
13
- "MatMul",
14
- "Add",
15
- "Div",
16
- "Gather",
17
- "Where",
18
- "Sigmoid",
19
  "ConstantOfShape",
20
- "Sub",
21
- "Cast",
22
- "Unsqueeze",
23
  "Transpose",
24
  "Expand",
25
- "Softmax",
26
- "ReduceMean",
27
- "Equal",
 
 
 
28
  "Reshape",
29
- "Concat"
 
 
 
 
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
  "decoder_with_past_model": {
34
  "op_types": [
35
- "Sqrt",
36
- "Shape",
37
  "Mul",
38
- "Pow",
39
- "Range",
40
  "Constant",
41
- "MatMul",
42
- "Add",
43
- "Div",
44
- "Gather",
45
- "Where",
46
- "Sigmoid",
47
  "ConstantOfShape",
48
- "Sub",
49
- "Cast",
50
- "Unsqueeze",
51
  "Transpose",
52
  "Expand",
53
- "Softmax",
54
- "ReduceMean",
55
- "Equal",
 
 
 
56
  "Reshape",
57
- "Concat"
 
 
 
 
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
63
- "Sqrt",
64
- "Shape",
65
  "Mul",
66
- "Pow",
67
- "Range",
68
- "Slice",
69
  "Constant",
70
- "MatMul",
71
- "Add",
72
- "Div",
73
- "Gather",
74
- "Where",
75
- "Sigmoid",
76
  "ConstantOfShape",
77
- "Sub",
 
78
  "Less",
79
- "Cast",
80
- "Unsqueeze",
81
  "Transpose",
82
- "Squeeze",
83
  "Expand",
84
- "Softmax",
85
- "ReduceMean",
86
- "Equal",
 
 
 
87
  "Reshape",
88
- "Concat"
 
 
 
 
 
 
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
94
- "Sqrt",
95
- "Shape",
96
  "Mul",
97
- "Pow",
98
- "Range",
99
- "Slice",
100
  "Constant",
101
- "MatMul",
102
- "Add",
103
- "Div",
104
- "Gather",
105
- "Where",
106
- "Sigmoid",
107
  "ConstantOfShape",
108
- "Sub",
 
109
  "Less",
110
- "Cast",
111
- "Unsqueeze",
112
  "Transpose",
113
- "Squeeze",
114
  "Expand",
115
- "Softmax",
116
- "ReduceMean",
117
- "Equal",
118
- "If",
 
 
119
  "Reshape",
120
- "Concat"
 
 
 
 
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
 
4
  "per_model_config": {
5
  "encoder_model": {
6
  "op_types": [
7
+ "Concat",
 
8
  "Mul",
9
+ "Shape",
 
10
  "Constant",
11
+ "Softmax",
12
+ "Equal",
 
 
 
 
13
  "ConstantOfShape",
14
+ "Pow",
15
+ "Where",
 
16
  "Transpose",
17
  "Expand",
18
+ "Sqrt",
19
+ "Gather",
20
+ "Div",
21
+ "Cast",
22
+ "Sigmoid",
23
+ "Range",
24
  "Reshape",
25
+ "Unsqueeze",
26
+ "Add",
27
+ "ReduceMean",
28
+ "MatMul",
29
+ "Sub"
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
  "decoder_with_past_model": {
34
  "op_types": [
35
+ "Concat",
 
36
  "Mul",
37
+ "Shape",
 
38
  "Constant",
39
+ "Softmax",
40
+ "Equal",
 
 
 
 
41
  "ConstantOfShape",
42
+ "Pow",
43
+ "Where",
 
44
  "Transpose",
45
  "Expand",
46
+ "Sqrt",
47
+ "Gather",
48
+ "Div",
49
+ "Cast",
50
+ "Sigmoid",
51
+ "Range",
52
  "Reshape",
53
+ "Unsqueeze",
54
+ "Add",
55
+ "ReduceMean",
56
+ "MatMul",
57
+ "Sub"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
+ "decoder_model_merged": {
62
  "op_types": [
63
+ "Concat",
 
64
  "Mul",
65
+ "Shape",
 
 
66
  "Constant",
67
+ "Softmax",
68
+ "Equal",
 
 
 
 
69
  "ConstantOfShape",
70
+ "Squeeze",
71
+ "Pow",
72
  "Less",
73
+ "Where",
 
74
  "Transpose",
 
75
  "Expand",
76
+ "Sqrt",
77
+ "Gather",
78
+ "Div",
79
+ "Cast",
80
+ "Sigmoid",
81
+ "Range",
82
  "Reshape",
83
+ "Unsqueeze",
84
+ "Add",
85
+ "If",
86
+ "ReduceMean",
87
+ "MatMul",
88
+ "Slice",
89
+ "Sub"
90
  ],
91
  "weight_type": "QInt8"
92
  },
93
+ "decoder_model": {
94
  "op_types": [
95
+ "Concat",
 
96
  "Mul",
97
+ "Shape",
 
 
98
  "Constant",
99
+ "Softmax",
100
+ "Equal",
 
 
 
 
101
  "ConstantOfShape",
102
+ "Squeeze",
103
+ "Pow",
104
  "Less",
105
+ "Where",
 
106
  "Transpose",
 
107
  "Expand",
108
+ "Sqrt",
109
+ "Gather",
110
+ "Div",
111
+ "Cast",
112
+ "Sigmoid",
113
+ "Range",
114
  "Reshape",
115
+ "Unsqueeze",
116
+ "Add",
117
+ "ReduceMean",
118
+ "MatMul",
119
+ "Slice",
120
+ "Sub"
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff