Xenova HF staff commited on
Commit
96fa762
1 Parent(s): 263e182

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -3,6 +3,7 @@
3
  "architectures": [
4
  "T5EncoderModel"
5
  ],
 
6
  "d_ff": 4096,
7
  "d_kv": 64,
8
  "d_model": 1024,
@@ -53,7 +54,7 @@
53
  "prefix": "translate English to Romanian: "
54
  }
55
  },
56
- "transformers_version": "4.30.2",
57
  "use_cache": true,
58
  "vocab_size": 32128
59
  }
 
3
  "architectures": [
4
  "T5EncoderModel"
5
  ],
6
+ "classifier_dropout": 0.0,
7
  "d_ff": 4096,
8
  "d_kv": 64,
9
  "d_model": 1024,
 
54
  "prefix": "translate English to Romanian: "
55
  }
56
  },
57
+ "transformers_version": "4.34.0.dev0",
58
  "use_cache": true,
59
  "vocab_size": 32128
60
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e575acafa28f7f35d90dcdd89ea649bbc8512bca0a1db79406dfdd6ee24f6278
3
  size 1742842287
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35f927ec83849d92c9b12beae10059da0a0b5e7e7d0d6c14d47578fc5a6ea0a5
3
  size 1742842287
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:51de4cf51502b431bf5d79503cae1c6c82f0a66e53e570342d3d1dca8df905a7
3
  size 1743425095
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac1c489a7fce2f22672b05806971a8065c97dd8141f85aaacb7e0064404fd32d
3
  size 1743425095
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4773559d4f7e3fcee3d347948088acbd2d1080bdc9bd253bb11858aa4f19350f
3
- size 439300935
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08781657cbbd53c53c074a90b18a320e89165d8fd5b4e1f098569e97b308eaa3
3
+ size 439300934
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:67ae9dd18c567b135ac307465b92185b8b3ac185cd7c5032bbd7a5778beb9a95
3
- size 438286627
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db8c665a063733beb2e3c6349d7a99673dd17f63d80a4deea4273ff67421bf90
3
+ size 438286626
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ffbb4a98b45af9fcd080cc26dae636a5185b39ceac3f8ff5263fa6caf48c2c5f
3
  size 1541451703
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc5e930426c9a370e7144f9249f86a6cf7feb15731792b266a56888bfd8c2220
3
  size 1541451703
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0ec6929d182573cb2a79ff9360289d985e8a050d4d40c0058f46d41b39f10359
3
- size 387577779
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58e543a43efabf3dd05e6d216037b5a82cf7a775eed8a82b3e0d8548801dad83
3
+ size 387577778
quantize_config.json ADDED
@@ -0,0 +1,142 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": true,
3
+ "reduce_range": true,
4
+ "per_model_config": {
5
+ "decoder_model": {
6
+ "op_types": [
7
+ "Unsqueeze",
8
+ "Reshape",
9
+ "Div",
10
+ "Mul",
11
+ "Identity",
12
+ "Where",
13
+ "Sub",
14
+ "ReduceMean",
15
+ "Tile",
16
+ "Log",
17
+ "LessOrEqual",
18
+ "Range",
19
+ "MatMul",
20
+ "Sqrt",
21
+ "Add",
22
+ "Transpose",
23
+ "Neg",
24
+ "Cast",
25
+ "Gather",
26
+ "Softmax",
27
+ "ConstantOfShape",
28
+ "Constant",
29
+ "Expand",
30
+ "Pow",
31
+ "Concat",
32
+ "Less",
33
+ "Min",
34
+ "Shape",
35
+ "Relu"
36
+ ],
37
+ "weight_type": "QInt8"
38
+ },
39
+ "encoder_model": {
40
+ "op_types": [
41
+ "Unsqueeze",
42
+ "Reshape",
43
+ "Div",
44
+ "Greater",
45
+ "Mul",
46
+ "Where",
47
+ "Abs",
48
+ "Sub",
49
+ "ReduceMean",
50
+ "Log",
51
+ "Range",
52
+ "MatMul",
53
+ "Sqrt",
54
+ "Add",
55
+ "Transpose",
56
+ "Cast",
57
+ "Gather",
58
+ "Softmax",
59
+ "ConstantOfShape",
60
+ "Constant",
61
+ "Pow",
62
+ "Concat",
63
+ "Less",
64
+ "Min",
65
+ "Shape",
66
+ "Relu"
67
+ ],
68
+ "weight_type": "QInt8"
69
+ },
70
+ "decoder_with_past_model": {
71
+ "op_types": [
72
+ "Unsqueeze",
73
+ "Reshape",
74
+ "Div",
75
+ "Slice",
76
+ "Mul",
77
+ "Identity",
78
+ "Where",
79
+ "Sub",
80
+ "ReduceMean",
81
+ "Tile",
82
+ "Log",
83
+ "LessOrEqual",
84
+ "Range",
85
+ "MatMul",
86
+ "Sqrt",
87
+ "Add",
88
+ "Transpose",
89
+ "Neg",
90
+ "Cast",
91
+ "Gather",
92
+ "Softmax",
93
+ "ConstantOfShape",
94
+ "Constant",
95
+ "Expand",
96
+ "Pow",
97
+ "Concat",
98
+ "Less",
99
+ "Min",
100
+ "Shape",
101
+ "Relu"
102
+ ],
103
+ "weight_type": "QInt8"
104
+ },
105
+ "decoder_model_merged": {
106
+ "op_types": [
107
+ "Unsqueeze",
108
+ "Reshape",
109
+ "Div",
110
+ "Slice",
111
+ "Mul",
112
+ "Identity",
113
+ "Where",
114
+ "Sub",
115
+ "ReduceMean",
116
+ "Tile",
117
+ "Log",
118
+ "LessOrEqual",
119
+ "Range",
120
+ "MatMul",
121
+ "Sqrt",
122
+ "Add",
123
+ "If",
124
+ "Transpose",
125
+ "Neg",
126
+ "Cast",
127
+ "Gather",
128
+ "Softmax",
129
+ "ConstantOfShape",
130
+ "Constant",
131
+ "Expand",
132
+ "Pow",
133
+ "Concat",
134
+ "Less",
135
+ "Min",
136
+ "Shape",
137
+ "Relu"
138
+ ],
139
+ "weight_type": "QInt8"
140
+ }
141
+ }
142
+ }
tokenizer_config.json CHANGED
@@ -104,8 +104,12 @@
104
  "clean_up_tokenization_spaces": true,
105
  "eos_token": "</s>",
106
  "extra_ids": 100,
 
107
  "model_max_length": 512,
 
108
  "pad_token": "<pad>",
 
 
109
  "tokenizer_class": "T5Tokenizer",
110
  "unk_token": "<unk>"
111
  }
 
104
  "clean_up_tokenization_spaces": true,
105
  "eos_token": "</s>",
106
  "extra_ids": 100,
107
+ "max_length": null,
108
  "model_max_length": 512,
109
+ "pad_to_multiple_of": null,
110
  "pad_token": "<pad>",
111
+ "pad_token_type_id": 0,
112
+ "padding_side": "right",
113
  "tokenizer_class": "T5Tokenizer",
114
  "unk_token": "<unk>"
115
  }