Xenova HF staff commited on
Commit
8f725e8
1 Parent(s): 87f31f0

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": 59514
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": 59514
61
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  59513
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 4,
14
  "pad_token_id": 59513,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  59513
 
11
  "max_length": 512,
12
  "num_beams": 4,
13
  "pad_token_id": 59513,
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:e02694246ed3ee7ecd1fbabbff83b0e496dd0577b603a32b0f0405768597126f
3
- size 346228960
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1844d3c6f595df70cd0745084afa8fd9e98227bce7ac8fb9a3f1ddd41bcd171b
3
+ size 224344345
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:05574d689c5d7138c6bc5572a0ac6dfd7ebc6c76b23102332885b6487c312a26
3
- size 346464442
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1981b3cddcfdc8783e1d201a9316a25202c504f8566a970960816e1dbe3d9c34
3
+ size 224579912
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9643c39f502bd8f124e1860b1ce4e09efbad91477b9f654c55f8595ed6dcda9a
3
- size 88151662
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73bc7ac8e29c42e6f212ebcc29a2991d3646a04aabb642c0036aa54b40e4e1a9
3
+ size 57381512
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:99e34db3e10d0886f2a6036cb772cd88082ea4768e2dcf546c866f89706aa1c7
3
- size 87780306
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c5a0c5d113a1c4c5ae46e014aacf770f3bff2109fe8b74ead6aae5c5f408ffc
3
+ size 57010810
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e5ce3bfea94b6a2dc5c293b313898e2e1017bb72dae80a4ead10b2385df3be54
3
- size 333578451
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7732e12e6cf0394be29432294b2df9dbd64cd1d56527d537acc209bd59bddbc7
3
+ size 211693836
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e73d33ad286cb15f7eddbc629ffdc6ed61f14823cefb125f4aa2e10fee548970
3
- size 84508842
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5052aa36fd6e4e10b53db8f59e21ac85d4ae7a8857e10575a31485c43a1583ad
3
+ size 53739346
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
- "Unsqueeze",
8
- "Sub",
9
- "Div",
10
- "Range",
11
- "Cast",
12
- "Equal",
13
- "Reshape",
14
  "Concat",
15
- "Shape",
16
- "Gather",
17
- "Where",
18
- "Constant",
19
- "MatMul",
20
- "Mul",
21
- "Sigmoid",
22
  "Softmax",
23
- "Transpose",
 
 
 
 
 
24
  "ConstantOfShape",
25
- "Expand",
 
 
 
26
  "Pow",
27
- "Add",
28
  "ReduceMean",
29
- "Sqrt"
 
 
 
 
 
 
 
 
 
 
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
- "decoder_with_past_model": {
34
  "op_types": [
35
- "Unsqueeze",
36
- "Sub",
37
- "Div",
38
- "Range",
39
- "Cast",
40
- "Equal",
41
- "Reshape",
42
  "Concat",
43
- "Shape",
44
- "Gather",
45
- "Where",
46
- "Constant",
47
- "MatMul",
48
- "Mul",
49
- "Sigmoid",
50
  "Softmax",
51
- "Transpose",
 
 
 
 
52
  "ConstantOfShape",
53
- "Expand",
 
 
54
  "Pow",
55
- "Add",
56
  "ReduceMean",
57
- "Sqrt"
 
 
 
 
 
 
 
 
 
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
63
- "Unsqueeze",
64
- "Sub",
65
- "Div",
66
- "Range",
67
  "Cast",
68
- "Slice",
69
- "Equal",
70
  "Squeeze",
71
- "Reshape",
72
- "Concat",
73
- "Shape",
74
  "Gather",
75
- "Where",
76
- "Constant",
77
- "MatMul",
78
- "Mul",
79
- "Sigmoid",
80
- "Softmax",
81
- "Transpose",
82
  "ConstantOfShape",
83
- "Expand",
 
 
 
84
  "Pow",
85
- "Add",
86
  "ReduceMean",
 
 
 
 
 
 
 
 
87
  "Sqrt",
88
- "Less"
 
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
94
- "Unsqueeze",
95
- "Sub",
96
- "Div",
97
- "Range",
98
- "Cast",
99
- "Slice",
100
- "Equal",
101
- "Squeeze",
102
- "Reshape",
103
- "Shape",
104
  "Concat",
105
- "Gather",
106
- "Where",
107
- "Constant",
108
- "MatMul",
109
- "Mul",
110
- "Sigmoid",
111
  "Softmax",
112
- "Transpose",
113
- "If",
 
 
 
114
  "ConstantOfShape",
115
- "Expand",
 
 
116
  "Pow",
117
- "Add",
118
  "ReduceMean",
 
 
 
 
 
 
 
119
  "Sqrt",
120
- "Less"
 
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
  "Concat",
 
 
 
 
 
 
 
8
  "Softmax",
9
+ "Cast",
10
+ "Where",
11
+ "Squeeze",
12
+ "Range",
13
+ "Gather",
14
+ "Add",
15
  "ConstantOfShape",
16
+ "Sigmoid",
17
+ "Reshape",
18
+ "Mul",
19
+ "Slice",
20
  "Pow",
 
21
  "ReduceMean",
22
+ "Less",
23
+ "Expand",
24
+ "Transpose",
25
+ "MatMul",
26
+ "Constant",
27
+ "Sub",
28
+ "Unsqueeze",
29
+ "Shape",
30
+ "Sqrt",
31
+ "Equal",
32
+ "Div"
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "encoder_model": {
37
  "op_types": [
 
 
 
 
 
 
 
38
  "Concat",
 
 
 
 
 
 
 
39
  "Softmax",
40
+ "Cast",
41
+ "Where",
42
+ "Range",
43
+ "Gather",
44
+ "Add",
45
  "ConstantOfShape",
46
+ "Sigmoid",
47
+ "Reshape",
48
+ "Mul",
49
  "Pow",
 
50
  "ReduceMean",
51
+ "Expand",
52
+ "Transpose",
53
+ "MatMul",
54
+ "Constant",
55
+ "Sub",
56
+ "Unsqueeze",
57
+ "Shape",
58
+ "Sqrt",
59
+ "Equal",
60
+ "Div"
61
  ],
62
  "weight_type": "QInt8"
63
  },
64
+ "decoder_model_merged": {
65
  "op_types": [
66
+ "Concat",
67
+ "If",
68
+ "Softmax",
 
69
  "Cast",
70
+ "Where",
 
71
  "Squeeze",
72
+ "Range",
 
 
73
  "Gather",
74
+ "Add",
 
 
 
 
 
 
75
  "ConstantOfShape",
76
+ "Sigmoid",
77
+ "Reshape",
78
+ "Mul",
79
+ "Slice",
80
  "Pow",
 
81
  "ReduceMean",
82
+ "Less",
83
+ "Expand",
84
+ "Transpose",
85
+ "MatMul",
86
+ "Constant",
87
+ "Sub",
88
+ "Unsqueeze",
89
+ "Shape",
90
  "Sqrt",
91
+ "Equal",
92
+ "Div"
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
+ "decoder_with_past_model": {
97
  "op_types": [
 
 
 
 
 
 
 
 
 
 
98
  "Concat",
 
 
 
 
 
 
99
  "Softmax",
100
+ "Cast",
101
+ "Where",
102
+ "Range",
103
+ "Gather",
104
+ "Add",
105
  "ConstantOfShape",
106
+ "Sigmoid",
107
+ "Reshape",
108
+ "Mul",
109
  "Pow",
 
110
  "ReduceMean",
111
+ "Expand",
112
+ "Transpose",
113
+ "MatMul",
114
+ "Constant",
115
+ "Sub",
116
+ "Unsqueeze",
117
+ "Shape",
118
  "Sqrt",
119
+ "Equal",
120
+ "Div"
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff