Hugging Face's logo

Xenova
/
flan-t5-base

Text2Text Generation
Transformers.js
ONNX
t5
Model card Files Community
Xenova HF staff commited on
Commit
ddb2d21
•
1 Parent(s): 5046f8b

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": 2048,
7
  "d_kv": 64,
8
  "d_model": 768,
@@ -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": 2048,
8
  "d_kv": 64,
9
  "d_model": 768,
 
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:c95d12a21a1d9d2bcbe2c94d231fde1fdf19a1c39b4ee7980221d2c47427b785
3
- size 650849270
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd2dc690641be4851a21e4040c1d8e74c7d51486761730947de96df254bbd1c0
3
+ size 650848961
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3d318e88d9336eeaa68c4f4e2a3ec0049412e9a06087ee24ea502034335d6a7d
3
- size 651183578
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3028379bd19273868c6238287293e2e03b44db14ee279ad8a639be725e9b888
3
+ size 651182887
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8e895d08335e7736f48261128b073715ce070022e68e463ee86fba2d0375952
3
- size 163918689
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92efb296407104c5723ff0719fbcee97eb980b4e9e9b6d04fba7ef35f67f408c
3
+ size 164739632
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:15310a62f5bad73a06b7e6d14ff1c5fcd0fbcd981ffc3604a4d44af381514dd9
3
- size 163341718
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:da32795e448c704c40a9bfd52eb0aefb3cd976555d0747ed0227c6653bad0410
3
+ size 164163118
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:470b83ed21be0c6f6744c8d4e049355aa620e1d2528c82f869b1378f31c84deb
3
- size 594197616
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b47d2f266e175d42559a29c6330483769baca5b82042814e981402b49d4efcb
3
+ size 594197309
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:27418affbf6f8c9725530f61a57a6c98eee99a5550d0da1fb5eabec3b91f5d17
3
- size 149125021
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99dc3a63a08925384e5c63d7696e654d0d58c0749737c7409e98b29d68632b6a
3
+ size 149854152
onnx/encoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:217b494b1c5af6d1b5660d65053bc0d9881a8b6b3e0476783b6f2590af51b4e2
3
  size 438697388
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94c438df6ed8f9479f3b8ce43d4903cb4817d7c1e0d5c47b86a7a2ca9de621c9
3
  size 438697388
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:18fd9791ccb91c04137357e3a98575619f9f37a7e07175d4d8dadd285598e62f
3
- size 110025778
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea3f6bf594f5e4133311a59d54d3471ae2440bbe613bca398248c2bd32a5b6ec
3
+ size 110502358
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
+ "Mul",
8
+ "MatMul",
9
+ "Neg",
10
+ "LessOrEqual",
11
+ "Expand",
12
+ "Constant",
13
+ "Gather",
14
+ "Min",
15
+ "Where",
16
+ "Tanh",
17
+ "Reshape",
18
+ "Range",
19
+ "Less",
20
+ "Transpose",
21
+ "Shape",
22
+ "Log",
23
+ "ConstantOfShape",
24
+ "Sqrt",
25
+ "ReduceMean",
26
+ "Sub",
27
+ "Add",
28
+ "Unsqueeze",
29
+ "Div",
30
+ "Pow",
31
+ "Concat",
32
+ "Tile",
33
+ "Softmax",
34
+ "Cast"
35
+ ],
36
+ "weight_type": "QInt8"
37
+ },
38
+ "encoder_model": {
39
+ "op_types": [
40
+ "Mul",
41
+ "MatMul",
42
+ "Constant",
43
+ "Gather",
44
+ "Greater",
45
+ "Min",
46
+ "Abs",
47
+ "Where",
48
+ "Tanh",
49
+ "Reshape",
50
+ "Range",
51
+ "Less",
52
+ "Transpose",
53
+ "Shape",
54
+ "Log",
55
+ "Sqrt",
56
+ "ConstantOfShape",
57
+ "ReduceMean",
58
+ "Sub",
59
+ "Add",
60
+ "Unsqueeze",
61
+ "Div",
62
+ "Pow",
63
+ "Concat",
64
+ "Softmax",
65
+ "Cast"
66
+ ],
67
+ "weight_type": "QInt8"
68
+ },
69
+ "decoder_model_merged": {
70
+ "op_types": [
71
+ "Mul",
72
+ "MatMul",
73
+ "Neg",
74
+ "LessOrEqual",
75
+ "Expand",
76
+ "Constant",
77
+ "Gather",
78
+ "Slice",
79
+ "Min",
80
+ "Where",
81
+ "Tanh",
82
+ "Reshape",
83
+ "If",
84
+ "Range",
85
+ "Less",
86
+ "Transpose",
87
+ "Shape",
88
+ "Log",
89
+ "ConstantOfShape",
90
+ "Sqrt",
91
+ "ReduceMean",
92
+ "Sub",
93
+ "Add",
94
+ "Unsqueeze",
95
+ "Div",
96
+ "Pow",
97
+ "Concat",
98
+ "Tile",
99
+ "Softmax",
100
+ "Cast"
101
+ ],
102
+ "weight_type": "QInt8"
103
+ },
104
+ "decoder_with_past_model": {
105
+ "op_types": [
106
+ "Mul",
107
+ "MatMul",
108
+ "Neg",
109
+ "LessOrEqual",
110
+ "Expand",
111
+ "Constant",
112
+ "Gather",
113
+ "Slice",
114
+ "Min",
115
+ "Where",
116
+ "Tanh",
117
+ "Reshape",
118
+ "Range",
119
+ "Less",
120
+ "Transpose",
121
+ "Shape",
122
+ "Log",
123
+ "ConstantOfShape",
124
+ "Sqrt",
125
+ "ReduceMean",
126
+ "Add",
127
+ "Sub",
128
+ "Unsqueeze",
129
+ "Div",
130
+ "Pow",
131
+ "Concat",
132
+ "Tile",
133
+ "Softmax",
134
+ "Cast"
135
+ ],
136
+ "weight_type": "QInt8"
137
+ }
138
+ }
139
+ }