Xenova HF staff commited on
Commit
4161b65
1 Parent(s): 2207b5a

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -3,6 +3,7 @@
3
  "architectures": [
4
  "T5EncoderModel"
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": "4.30.2",
57
  "use_cache": true,
58
  "vocab_size": 32128
59
  }
 
3
  "architectures": [
4
  "T5EncoderModel"
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
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1c94f6ca0c9559f5226afed50b93da375594c033dacf4ebddb04de050cbb0179
3
  size 552007899
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5bcdfe706b32b65daea716dc47afe7117a8fcf933b2874ec8e164975940676d9
3
  size 552007899
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:96334aab0e1228e27832a44df67f33d1fb46f0a86f50a12105b29c9500eb08a1
3
  size 552310338
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b9657348ec7f0fa8b9fcbf28886321f16cc0b80aa63307210d385d574c50c6e
3
  size 552310338
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f0c09554009f8cddc37af1123fb2d713c35a109a5487cdeb3ce5c3881086c5b5
3
- size 139622447
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b48c92cdd2a243bc48bf7143f3277b4186639960123618ea820d6ad69af1ce78
3
+ size 139622448
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b7afa2207997de157633daafda3367ca22775606f26292dd8f0e4a16e7abc308
3
- size 139099971
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e562e2a181849dfa026c3b62fee9f47983ab184eb67d8170b9e194532c6a9295
3
+ size 139099972
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:50e0959664f8e0a84f7ec95ef5561b4b9f741f8f3164d80b6a1717caaeb62a68
3
  size 495356246
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bcaab450628a799758b9663af81f649e80a07ca7e09b76d0ab7fc302258eb092
3
  size 495356246
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b33e3d3ca76ff84d05539d3180296c37cb464c8ccaec73bb264d034ea68f4aff
3
- size 124791005
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b91961ebf68fbba8d0d1651129be239da6424df96108ea39a55a2c2b3bd48322
3
+ size 124791006
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
+ "Div",
8
+ "Min",
9
+ "MatMul",
10
+ "Expand",
11
+ "Pow",
12
+ "ConstantOfShape",
13
+ "Concat",
14
+ "LessOrEqual",
15
+ "Range",
16
+ "Constant",
17
+ "Shape",
18
+ "Where",
19
+ "Relu",
20
+ "Cast",
21
+ "Sqrt",
22
+ "Sub",
23
+ "Add",
24
+ "Log",
25
+ "Identity",
26
+ "ReduceMean",
27
+ "Unsqueeze",
28
+ "Neg",
29
+ "Mul",
30
+ "Softmax",
31
+ "Tile",
32
+ "Reshape",
33
+ "Less",
34
+ "Transpose",
35
+ "Gather"
36
+ ],
37
+ "weight_type": "QInt8"
38
+ },
39
+ "encoder_model": {
40
+ "op_types": [
41
+ "Div",
42
+ "Min",
43
+ "MatMul",
44
+ "Pow",
45
+ "ConstantOfShape",
46
+ "Concat",
47
+ "Range",
48
+ "Constant",
49
+ "Shape",
50
+ "Where",
51
+ "Relu",
52
+ "Cast",
53
+ "Sqrt",
54
+ "Sub",
55
+ "Add",
56
+ "Log",
57
+ "Abs",
58
+ "ReduceMean",
59
+ "Unsqueeze",
60
+ "Mul",
61
+ "Softmax",
62
+ "Reshape",
63
+ "Less",
64
+ "Transpose",
65
+ "Greater",
66
+ "Gather"
67
+ ],
68
+ "weight_type": "QInt8"
69
+ },
70
+ "decoder_with_past_model": {
71
+ "op_types": [
72
+ "Div",
73
+ "Min",
74
+ "MatMul",
75
+ "Expand",
76
+ "Pow",
77
+ "ConstantOfShape",
78
+ "Concat",
79
+ "LessOrEqual",
80
+ "Range",
81
+ "Constant",
82
+ "Shape",
83
+ "Where",
84
+ "Relu",
85
+ "Cast",
86
+ "Sqrt",
87
+ "Add",
88
+ "Sub",
89
+ "Slice",
90
+ "Log",
91
+ "Identity",
92
+ "ReduceMean",
93
+ "Unsqueeze",
94
+ "Neg",
95
+ "Mul",
96
+ "Softmax",
97
+ "Tile",
98
+ "Reshape",
99
+ "Less",
100
+ "Transpose",
101
+ "Gather"
102
+ ],
103
+ "weight_type": "QInt8"
104
+ },
105
+ "decoder_model_merged": {
106
+ "op_types": [
107
+ "Div",
108
+ "Min",
109
+ "MatMul",
110
+ "Expand",
111
+ "Pow",
112
+ "ConstantOfShape",
113
+ "Concat",
114
+ "LessOrEqual",
115
+ "Range",
116
+ "Constant",
117
+ "Shape",
118
+ "Where",
119
+ "Relu",
120
+ "Cast",
121
+ "Sqrt",
122
+ "Sub",
123
+ "Add",
124
+ "Slice",
125
+ "Log",
126
+ "Identity",
127
+ "ReduceMean",
128
+ "Unsqueeze",
129
+ "If",
130
+ "Neg",
131
+ "Mul",
132
+ "Softmax",
133
+ "Tile",
134
+ "Reshape",
135
+ "Less",
136
+ "Transpose",
137
+ "Gather"
138
+ ],
139
+ "weight_type": "QInt8"
140
+ }
141
+ }
142
+ }
tokenizer_config.json CHANGED
@@ -104,8 +104,15 @@
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": 512,
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
+ "stride": 0,
114
  "tokenizer_class": "T5Tokenizer",
115
+ "truncation_side": "right",
116
+ "truncation_strategy": "longest_first",
117
  "unk_token": "<unk>"
118
  }