Hugging Face's logo

Xenova
/
flan-t5-small

Text2Text Generation
Transformers.js
ONNX
t5
Model card Files Community
Xenova HF staff commited on
Commit
624f447
•
1 Parent(s): 2b7b966

Upload folder using huggingface_hub

Browse files
Files changed (11) 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. onnx/encoder_model.onnx +1 -1
  10. onnx/encoder_model_quantized.onnx +2 -2
  11. quantize_config.json +139 -0
config.json CHANGED
@@ -3,6 +3,7 @@
3
  "architectures": [
4
  "T5ForConditionalGeneration"
5
  ],
 
6
  "d_ff": 1024,
7
  "d_kv": 64,
8
  "d_model": 512,
@@ -54,7 +55,7 @@
54
  }
55
  },
56
  "tie_word_embeddings": false,
57
- "transformers_version": "4.28.1",
58
  "use_cache": true,
59
  "vocab_size": 32128
60
  }
 
3
  "architectures": [
4
  "T5ForConditionalGeneration"
5
  ],
6
+ "classifier_dropout": 0.0,
7
  "d_ff": 1024,
8
  "d_kv": 64,
9
  "d_model": 512,
 
55
  }
56
  },
57
  "tie_word_embeddings": false,
58
+ "transformers_version": "4.33.0.dev0",
59
  "use_cache": true,
60
  "vocab_size": 32128
61
  }
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": "4.28.1"
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 sha256:9a8e64ac0f50e81c2ad9b0c04720de0295a81fef90f02e1f44a803511be67453
3
- size 232553947
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76b9ffe1a3ee611c73bfba19a2fe4256e67261c7e4b8895546c8fa92001adf3a
3
+ size 232553640
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:de7e0d3873db023c50008ab64730ba70347c312a8ac4c7977e3cbb60cb16af9e
3
- size 232785013
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0bae0554dd4d1deaf6401bb2168ff7213167ffb877f6d7de4610d33e201f8142
3
+ size 232784326
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cc8bfd59455181bdc33e9d5949279cdd45b6a08e7d045d9a94b23993e017b5ef
3
- size 58943561
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73e2e942503221d7844715a8d824f68d8a2e878483f1c923849c177f0c441df1
3
+ size 59339331
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4f474acc3822860414709a0f192eb11c0abf8b4ee8fac5a6fa3bc0e786de7385
3
- size 58547095
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f090cbd1281a9c896a070b424ceb18df45da6e78d16895a14c72387708db216
3
+ size 58943318
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ac4e3d260196b56d01fb69a600899bd900a7c2e6595645aeb225aeea1cc0d213
3
- size 219954259
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9141efb7a5a7de265f21668729ddf77b74988dc130b3e8b86f0b7265fd27bfbc
3
+ size 219953954
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:64b0ef87d2e43e646d3c96411225c9d4f0912f9cdad56d253e9b102cb8983f07
3
- size 55363747
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:564669d8e39fbb8d9bf628070387e48901bfd8b8a9d76b0accd7ed637ef20a39
3
+ size 55729181
onnx/encoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2199dd4047cfddd8fa1c9927435e0a91a3b1f36dd01e21282bdff1348d95441c
3
  size 141456352
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07153ef2fd6eae4ccc29e579c66d3829ac40bbfc26fdc0b6a590cd3f55422b67
3
  size 141456352
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7843a74ae7606500d7799e8352025e4d4a745ed3814fd21de465e9454d443a85
3
- size 35589887
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:968ce399b8dcdfd4c36833db51e9f78f327f7272ee9e2777fc2b3c8693208e90
3
+ size 35759127
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
+ "LessOrEqual",
9
+ "ConstantOfShape",
10
+ "Expand",
11
+ "Pow",
12
+ "Less",
13
+ "Log",
14
+ "ReduceMean",
15
+ "Sqrt",
16
+ "Tanh",
17
+ "Unsqueeze",
18
+ "Min",
19
+ "Mul",
20
+ "Concat",
21
+ "Reshape",
22
+ "Gather",
23
+ "Where",
24
+ "Softmax",
25
+ "Div",
26
+ "Transpose",
27
+ "Cast",
28
+ "Shape",
29
+ "MatMul",
30
+ "Add",
31
+ "Sub",
32
+ "Constant",
33
+ "Neg",
34
+ "Range"
35
+ ],
36
+ "weight_type": "QInt8"
37
+ },
38
+ "encoder_model": {
39
+ "op_types": [
40
+ "Pow",
41
+ "ConstantOfShape",
42
+ "Less",
43
+ "Log",
44
+ "ReduceMean",
45
+ "Sqrt",
46
+ "Tanh",
47
+ "Unsqueeze",
48
+ "Min",
49
+ "Concat",
50
+ "Reshape",
51
+ "Gather",
52
+ "Where",
53
+ "Softmax",
54
+ "Div",
55
+ "Transpose",
56
+ "Cast",
57
+ "Shape",
58
+ "MatMul",
59
+ "Greater",
60
+ "Add",
61
+ "Range",
62
+ "Sub",
63
+ "Constant",
64
+ "Abs",
65
+ "Mul"
66
+ ],
67
+ "weight_type": "QInt8"
68
+ },
69
+ "decoder_model_merged": {
70
+ "op_types": [
71
+ "Tile",
72
+ "LessOrEqual",
73
+ "ConstantOfShape",
74
+ "Expand",
75
+ "Pow",
76
+ "Less",
77
+ "Log",
78
+ "ReduceMean",
79
+ "Sqrt",
80
+ "Tanh",
81
+ "Unsqueeze",
82
+ "Min",
83
+ "Mul",
84
+ "Concat",
85
+ "Reshape",
86
+ "Gather",
87
+ "If",
88
+ "Where",
89
+ "Softmax",
90
+ "Div",
91
+ "Slice",
92
+ "Transpose",
93
+ "Cast",
94
+ "Shape",
95
+ "MatMul",
96
+ "Add",
97
+ "Sub",
98
+ "Constant",
99
+ "Neg",
100
+ "Range"
101
+ ],
102
+ "weight_type": "QInt8"
103
+ },
104
+ "decoder_with_past_model": {
105
+ "op_types": [
106
+ "Tile",
107
+ "LessOrEqual",
108
+ "ConstantOfShape",
109
+ "Expand",
110
+ "Pow",
111
+ "Less",
112
+ "Log",
113
+ "ReduceMean",
114
+ "Sqrt",
115
+ "Tanh",
116
+ "Unsqueeze",
117
+ "Min",
118
+ "Mul",
119
+ "Concat",
120
+ "Reshape",
121
+ "Gather",
122
+ "Where",
123
+ "Softmax",
124
+ "Div",
125
+ "Slice",
126
+ "Transpose",
127
+ "Cast",
128
+ "Shape",
129
+ "MatMul",
130
+ "Add",
131
+ "Sub",
132
+ "Constant",
133
+ "Neg",
134
+ "Range"
135
+ ],
136
+ "weight_type": "QInt8"
137
+ }
138
+ }
139
+ }