Xenova HF staff commited on
Commit
ebeb069
1 Parent(s): f841c6a

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 sha256:0a452a1c374b2f1e670c38061355a8d569c0406b0eaaab8d873114b9fd9bbcd6
3
- size 650814243
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba2ad2d79c071d624b2ff940ad1fa4d5bb84746515143e6cc0027e31e7985a9d
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:cc550029785363b79975af201f326fec26a0099ea6c1baccaf5ce25c5c5b2258
3
- size 651113646
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0fa833a9e49f0d2ab22a5a42d93e85a4bd6a4f623d441a35ca8fed6362802efd
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:cebf87935e8b9c8782a448836e70e1bd53c0710ff0418b2fe8d67717a1b22a7c
3
- size 164561052
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d27a97155b87577ae89162a1d5f32b54cbde77cc21a53d1e97794d814957d6a2
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:b627f37dc41b9d8d00465dc8ab9b0109771133745701af8d136d7325f6d5d93e
3
- size 164042850
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1fc13b47da093ff92bbf043917aad498f28bd572635f3b3dc4f2a615f23d8477
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 sha256:11825b7d09480bab8d1c84131f850bbe7d14ed78e0f81e4615c0d401d1f7b887
3
- size 594162591
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc91b6053d9f8ad91c5c46a14369fa90b28841b2cfa4da8d6ba1687ff3a8a4d4
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 sha256:8cf58b24be099b90456c914a4c7bb07464d58856b0bb69a7dc775ef4026ebabf
3
- size 149733876
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53c18d2cc83b9cef64759037651ecf63089da944d15050ffc8704a43bfc7209d
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
+ "Gather",
8
+ "Add",
9
+ "Log",
10
+ "Unsqueeze",
11
+ "Reshape",
12
+ "LessOrEqual",
13
+ "ReduceMean",
14
+ "Transpose",
15
+ "Cast",
16
+ "Min",
17
+ "Tile",
18
+ "Concat",
19
+ "Sub",
20
+ "Div",
21
+ "Mul",
22
+ "Less",
23
+ "ConstantOfShape",
24
+ "Expand",
25
+ "Pow",
26
+ "MatMul",
27
+ "Shape",
28
+ "Sqrt",
29
+ "Relu",
30
+ "Where",
31
+ "Range",
32
+ "Neg",
33
+ "Softmax",
34
+ "Constant"
35
+ ],
36
+ "weight_type": "QInt8"
37
+ },
38
+ "encoder_model": {
39
+ "op_types": [
40
+ "Gather",
41
+ "Add",
42
+ "Log",
43
+ "Unsqueeze",
44
+ "Reshape",
45
+ "ReduceMean",
46
+ "Transpose",
47
+ "Cast",
48
+ "Min",
49
+ "Concat",
50
+ "Sub",
51
+ "Div",
52
+ "Mul",
53
+ "Less",
54
+ "ConstantOfShape",
55
+ "Pow",
56
+ "MatMul",
57
+ "Shape",
58
+ "Sqrt",
59
+ "Relu",
60
+ "Where",
61
+ "Range",
62
+ "Abs",
63
+ "Greater",
64
+ "Softmax",
65
+ "Constant"
66
+ ],
67
+ "weight_type": "QInt8"
68
+ },
69
+ "decoder_model_merged": {
70
+ "op_types": [
71
+ "Gather",
72
+ "Add",
73
+ "Log",
74
+ "Unsqueeze",
75
+ "Reshape",
76
+ "LessOrEqual",
77
+ "ReduceMean",
78
+ "Transpose",
79
+ "Cast",
80
+ "Min",
81
+ "If",
82
+ "Tile",
83
+ "Concat",
84
+ "Sub",
85
+ "Div",
86
+ "Mul",
87
+ "Less",
88
+ "ConstantOfShape",
89
+ "Expand",
90
+ "Pow",
91
+ "Slice",
92
+ "MatMul",
93
+ "Shape",
94
+ "Sqrt",
95
+ "Relu",
96
+ "Where",
97
+ "Range",
98
+ "Neg",
99
+ "Softmax",
100
+ "Constant"
101
+ ],
102
+ "weight_type": "QInt8"
103
+ },
104
+ "decoder_with_past_model": {
105
+ "op_types": [
106
+ "Gather",
107
+ "Add",
108
+ "Log",
109
+ "Unsqueeze",
110
+ "Reshape",
111
+ "LessOrEqual",
112
+ "ReduceMean",
113
+ "Transpose",
114
+ "Cast",
115
+ "Min",
116
+ "Tile",
117
+ "Concat",
118
+ "Sub",
119
+ "Div",
120
+ "Mul",
121
+ "Less",
122
+ "ConstantOfShape",
123
+ "Expand",
124
+ "Pow",
125
+ "Slice",
126
+ "MatMul",
127
+ "Shape",
128
+ "Sqrt",
129
+ "Relu",
130
+ "Where",
131
+ "Range",
132
+ "Neg",
133
+ "Softmax",
134
+ "Constant"
135
+ ],
136
+ "weight_type": "QInt8"
137
+ }
138
+ }
139
+ }