Hugging Face's logo

Xenova
/
mt5-small

Text2Text Generation
Transformers.js
ONNX
mt5
Model card Files Community
1
Xenova HF Staff commited on
Commit
7edab8f
·
1 Parent(s): 891b18d

Upload folder using huggingface_hub

Browse files
Files changed (10) hide show
  1. config.json +2 -1
  2. generation_config.json +1 -1
  3. onnx/decoder_model.onnx +2 -2
  4. onnx/decoder_model_merged.onnx +2 -2
  5. onnx/decoder_model_merged_quantized.onnx +2 -2
  6. onnx/decoder_model_quantized.onnx +2 -2
  7. onnx/decoder_with_past_model.onnx +2 -2
  8. onnx/decoder_with_past_model_quantized.onnx +2 -2
  9. quantize_config.json +139 -0
  10. tokenizer_config.json +1 -0
config.json CHANGED
@@ -3,6 +3,7 @@
3
  "architectures": [
4
  "MT5ForConditionalGeneration"
5
  ],
 
6
  "d_ff": 1024,
7
  "d_kv": 64,
8
  "d_model": 512,
@@ -24,7 +25,7 @@
24
  "relative_attention_num_buckets": 32,
25
  "tie_word_embeddings": false,
26
  "tokenizer_class": "T5Tokenizer",
27
- "transformers_version": "4.29.2",
28
  "use_cache": true,
29
  "vocab_size": 250112
30
  }
 
3
  "architectures": [
4
  "MT5ForConditionalGeneration"
5
  ],
6
+ "classifier_dropout": 0.0,
7
  "d_ff": 1024,
8
  "d_kv": 64,
9
  "d_model": 512,
 
25
  "relative_attention_num_buckets": 32,
26
  "tie_word_embeddings": false,
27
  "tokenizer_class": "T5Tokenizer",
28
+ "transformers_version": "4.33.0.dev0",
29
  "use_cache": true,
30
  "vocab_size": 250112
31
  }
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": "4.33.0.dev0"
7
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:58fb09e90e612342c6134ea6cfb3d373a0c0a3830d0151b62374342e9b19d338
3
- size 1125416416
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:701b93c03fdcc3c243fd7a8be4b3373eaf80a4b7878ab5f6472b34fc6faf4ea5
3
+ size 1125416109
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9d575c570aa9cee2a91c4dfe39df28236ef7b1bc78a24c665d409953bdaf3dd6
3
- size 1125647482
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37907b00bbf37057c8a6443d02461dd9330db268beb2233ad113a65a6fe23d93
3
+ size 1125646795
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e11ff57dbcb4e64ff79786549502179af3591cc7f90eba50e3e5cfee4f186492
3
- size 283645709
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61b2e843dd174874268cccd17aa1f6445bcd739af98507b2ac4758bd81c1c9be
3
+ size 283644868
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:37e433df2928d430614db560b90c989c568374e776b3767fabadf409c0d5757c
3
- size 283249243
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69e0a3e53dad4cf0702233e991490d3f359188122e20bbf04b67244d15a042b4
3
+ size 283248855
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e0c8eb8670ff73978a1e048da89bbd5acab340ae7b5e002d69664b779e4a63db
3
- size 1112816728
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:755bcd2d76017e69dea38bb864d7d2a0d0f3808481dd57206907bd411c41516e
3
+ size 1112816423
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a949cad5bf7ae4f7bf9163e2b509991979914d036ff5ada8a11bd0fadf2a02eb
3
- size 280035095
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89c588bf6fa8b168a6d2c780a854bc192ff994ae6a65006946415abd372410af
3
+ size 280034718
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
+ "Tile",
8
+ "MatMul",
9
+ "ConstantOfShape",
10
+ "Sub",
11
+ "Expand",
12
+ "Cast",
13
+ "LessOrEqual",
14
+ "Log",
15
+ "Sqrt",
16
+ "Less",
17
+ "Tanh",
18
+ "Shape",
19
+ "Neg",
20
+ "Mul",
21
+ "Softmax",
22
+ "Constant",
23
+ "Pow",
24
+ "Reshape",
25
+ "ReduceMean",
26
+ "Gather",
27
+ "Add",
28
+ "Where",
29
+ "Transpose",
30
+ "Concat",
31
+ "Range",
32
+ "Div",
33
+ "Unsqueeze",
34
+ "Min"
35
+ ],
36
+ "weight_type": "QInt8"
37
+ },
38
+ "encoder_model": {
39
+ "op_types": [
40
+ "MatMul",
41
+ "ConstantOfShape",
42
+ "Sub",
43
+ "Cast",
44
+ "Log",
45
+ "Sqrt",
46
+ "Less",
47
+ "Tanh",
48
+ "Shape",
49
+ "Mul",
50
+ "Softmax",
51
+ "Constant",
52
+ "Pow",
53
+ "Reshape",
54
+ "ReduceMean",
55
+ "Gather",
56
+ "Add",
57
+ "Where",
58
+ "Transpose",
59
+ "Concat",
60
+ "Range",
61
+ "Div",
62
+ "Unsqueeze",
63
+ "Min",
64
+ "Abs",
65
+ "Greater"
66
+ ],
67
+ "weight_type": "QInt8"
68
+ },
69
+ "decoder_model_merged": {
70
+ "op_types": [
71
+ "Tile",
72
+ "MatMul",
73
+ "ConstantOfShape",
74
+ "Sub",
75
+ "Expand",
76
+ "Cast",
77
+ "LessOrEqual",
78
+ "Log",
79
+ "Sqrt",
80
+ "Less",
81
+ "Tanh",
82
+ "Shape",
83
+ "Neg",
84
+ "Mul",
85
+ "If",
86
+ "Softmax",
87
+ "Constant",
88
+ "Slice",
89
+ "Pow",
90
+ "Reshape",
91
+ "ReduceMean",
92
+ "Gather",
93
+ "Add",
94
+ "Where",
95
+ "Transpose",
96
+ "Concat",
97
+ "Range",
98
+ "Div",
99
+ "Unsqueeze",
100
+ "Min"
101
+ ],
102
+ "weight_type": "QInt8"
103
+ },
104
+ "decoder_with_past_model": {
105
+ "op_types": [
106
+ "Tile",
107
+ "MatMul",
108
+ "ConstantOfShape",
109
+ "Sub",
110
+ "Expand",
111
+ "Cast",
112
+ "LessOrEqual",
113
+ "Log",
114
+ "Sqrt",
115
+ "Less",
116
+ "Tanh",
117
+ "Shape",
118
+ "Neg",
119
+ "Mul",
120
+ "Softmax",
121
+ "Constant",
122
+ "Slice",
123
+ "Pow",
124
+ "Reshape",
125
+ "ReduceMean",
126
+ "Gather",
127
+ "Add",
128
+ "Where",
129
+ "Transpose",
130
+ "Concat",
131
+ "Range",
132
+ "Div",
133
+ "Unsqueeze",
134
+ "Min"
135
+ ],
136
+ "weight_type": "QInt8"
137
+ }
138
+ }
139
+ }
tokenizer_config.json CHANGED
@@ -3,6 +3,7 @@
3
  "clean_up_tokenization_spaces": true,
4
  "eos_token": "</s>",
5
  "extra_ids": 0,
 
6
  "model_max_length": 1000000000000000019884624838656,
7
  "pad_token": "<pad>",
8
  "sp_model_kwargs": {},
 
3
  "clean_up_tokenization_spaces": true,
4
  "eos_token": "</s>",
5
  "extra_ids": 0,
6
+ "legacy": true,
7
  "model_max_length": 1000000000000000019884624838656,
8
  "pad_token": "<pad>",
9
  "sp_model_kwargs": {},