Xenova HF staff commited on
Commit
bda1337
1 Parent(s): e0941c5

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -3,6 +3,7 @@
3
  "architectures": [
4
  "T5ForConditionalGeneration"
5
  ],
 
6
  "d_ff": 3072,
7
  "d_kv": 64,
8
  "d_model": 768,
@@ -53,7 +54,7 @@
53
  "prefix": "translate English to Romanian: "
54
  }
55
  },
56
- "transformers_version": "4.29.2",
57
  "use_cache": true,
58
  "vocab_size": 32128
59
  }
 
3
  "architectures": [
4
  "T5ForConditionalGeneration"
5
  ],
6
+ "classifier_dropout": 0.0,
7
  "d_ff": 3072,
8
  "d_kv": 64,
9
  "d_model": 768,
 
54
  "prefix": "translate English to Romanian: "
55
  }
56
  },
57
+ "transformers_version": "4.34.0.dev0",
58
  "use_cache": true,
59
  "vocab_size": 32128
60
  }
generation_config.json CHANGED
@@ -3,5 +3,5 @@
3
  "decoder_start_token_id": 0,
4
  "eos_token_id": 1,
5
  "pad_token_id": 0,
6
- "transformers_version": "4.29.2"
7
  }
 
3
  "decoder_start_token_id": 0,
4
  "eos_token_id": 1,
5
  "pad_token_id": 0,
6
+ "transformers_version": "4.34.0.dev0"
7
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7ee62138e66e199d123cf56dc8b5590245028cdaa55708bd0a986de89dcbcc86
3
- size 650814243
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6242369f92d2f5aa7cbe26d12cd2cb364e1bde456f2c28bd08eaa5d94ef50c87
3
+ size 552116757
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:47dbd8e1f54dc9af53effa8cad4993269773f272092d8505631aae6720219d4d
3
- size 651113646
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ffd5ddee6a46abee218d33feebd648ebf24c09040ebb4f25d86e6a344d760a7
3
+ size 552415846
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2cf725d1960ccb9ab09a2a6a1fcec3d1549cfbc39ba5ded56807ba3d4739cff8
3
- size 164561052
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2494ce05c8e9956bb9862e6e37b83632e5c6f0d8489f336e5826b558072cb915
3
+ size 139724510
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a146e1cb900e471b79b41033e4d30ebed6c16bd49d188cd3436b71e1e7b40669
3
- size 164042850
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f76ab563bb9b6682417537dc1f2cc2737607f8ab7c3ac400626f1fee0d6c6f2
3
+ size 139207089
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:63d5beac5fe031d2d2a009309a8a11e4b0942ef799882fc9527242f6301a99ae
3
- size 594162591
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1ee7f2ff3b96941baabc6be486c77442edb35aa05217c620919ef6cb3945be0
3
+ size 495465107
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c00b0599645e0e0b3fbe24f55bf1925c93fa9ad2455eb4e7e3c96a7d891153d0
3
- size 149733876
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e7a0547232b8722295f533d3fc2e4bb7046a11d1660f25c0279a33b707e4901
3
+ size 124898126
quantize_config.json ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": true,
3
+ "reduce_range": true,
4
+ "per_model_config": {
5
+ "decoder_model": {
6
+ "op_types": [
7
+ "Add",
8
+ "Shape",
9
+ "Pow",
10
+ "Gather",
11
+ "Where",
12
+ "Mul",
13
+ "ReduceMean",
14
+ "Concat",
15
+ "ConstantOfShape",
16
+ "Less",
17
+ "Tile",
18
+ "Cast",
19
+ "MatMul",
20
+ "Unsqueeze",
21
+ "LessOrEqual",
22
+ "Transpose",
23
+ "Min",
24
+ "Reshape",
25
+ "Expand",
26
+ "Range",
27
+ "Constant",
28
+ "Log",
29
+ "Neg",
30
+ "Softmax",
31
+ "Sqrt",
32
+ "Relu",
33
+ "Div",
34
+ "Sub"
35
+ ],
36
+ "weight_type": "QInt8"
37
+ },
38
+ "encoder_model": {
39
+ "op_types": [
40
+ "Abs",
41
+ "Add",
42
+ "Shape",
43
+ "Pow",
44
+ "Gather",
45
+ "Where",
46
+ "Mul",
47
+ "ReduceMean",
48
+ "Concat",
49
+ "ConstantOfShape",
50
+ "Less",
51
+ "MatMul",
52
+ "Cast",
53
+ "Unsqueeze",
54
+ "Greater",
55
+ "Transpose",
56
+ "Log",
57
+ "Reshape",
58
+ "Min",
59
+ "Range",
60
+ "Constant",
61
+ "Softmax",
62
+ "Sqrt",
63
+ "Relu",
64
+ "Div",
65
+ "Sub"
66
+ ],
67
+ "weight_type": "QInt8"
68
+ },
69
+ "decoder_model_merged": {
70
+ "op_types": [
71
+ "Add",
72
+ "Shape",
73
+ "Pow",
74
+ "Gather",
75
+ "Where",
76
+ "Mul",
77
+ "ReduceMean",
78
+ "Concat",
79
+ "ConstantOfShape",
80
+ "Less",
81
+ "Tile",
82
+ "Cast",
83
+ "MatMul",
84
+ "Unsqueeze",
85
+ "Slice",
86
+ "LessOrEqual",
87
+ "Transpose",
88
+ "Min",
89
+ "Reshape",
90
+ "Expand",
91
+ "If",
92
+ "Range",
93
+ "Constant",
94
+ "Neg",
95
+ "Log",
96
+ "Softmax",
97
+ "Sqrt",
98
+ "Relu",
99
+ "Div",
100
+ "Sub"
101
+ ],
102
+ "weight_type": "QInt8"
103
+ },
104
+ "decoder_with_past_model": {
105
+ "op_types": [
106
+ "Add",
107
+ "Shape",
108
+ "Pow",
109
+ "Gather",
110
+ "Where",
111
+ "Mul",
112
+ "ReduceMean",
113
+ "Concat",
114
+ "ConstantOfShape",
115
+ "Less",
116
+ "Tile",
117
+ "Cast",
118
+ "MatMul",
119
+ "Unsqueeze",
120
+ "Slice",
121
+ "LessOrEqual",
122
+ "Transpose",
123
+ "Min",
124
+ "Reshape",
125
+ "Expand",
126
+ "Range",
127
+ "Constant",
128
+ "Log",
129
+ "Neg",
130
+ "Softmax",
131
+ "Sqrt",
132
+ "Relu",
133
+ "Div",
134
+ "Sub"
135
+ ],
136
+ "weight_type": "QInt8"
137
+ }
138
+ }
139
+ }