Hugging Face's logo

Xenova
/
LaMini-Flan-T5-783M

Transformers.js
ONNX
t5
text2text-generation
Model card Files
xet
Community
4
Xenova HF Staff commited on
Commit
24675c9
·
1 Parent(s): 095365e

Upload folder using huggingface_hub

Browse files
Files changed (11) hide show
  1. config.json +2 -2
  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": 2816,
7
  "d_kv": 64,
8
  "d_model": 1024,
@@ -25,8 +26,7 @@
25
  "relative_attention_max_distance": 128,
26
  "relative_attention_num_buckets": 32,
27
  "tie_word_embeddings": false,
28
- "torch_dtype": "float32",
29
- "transformers_version": "4.28.1",
30
  "use_cache": true,
31
  "vocab_size": 32128
32
  }
 
3
  "architectures": [
4
  "T5ForConditionalGeneration"
5
  ],
6
+ "classifier_dropout": 0.0,
7
  "d_ff": 2816,
8
  "d_kv": 64,
9
  "d_model": 1024,
 
26
  "relative_attention_max_distance": 128,
27
  "relative_attention_num_buckets": 32,
28
  "tie_word_embeddings": false,
29
+ "transformers_version": "4.34.0.dev0",
 
30
  "use_cache": true,
31
  "vocab_size": 32128
32
  }
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": "4.34.0.dev0"
7
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:643891f8024e09fcaaf21c7400aea149d8cc88595c2c585db5f82efb4619ce25
3
- size 1899966406
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0af2d5176759dac258e16aa9b5421584e608a20bb8d1234871df44bf7cf2ef05
3
+ size 1899966097
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4f8428ae370f32d9abf39c3353c390bb40889b352f40bbb3da5da09dc5330529
3
- size 1900612838
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d5e34da7b389b73d0aa02aaf5e8e6d42c1299294f77f116772cdf36686d25c2
3
+ size 1900612149
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f01eaef9c2c32baab84ec9101c67b1371a74cd9b437cfa71156dfef6419449ff
3
- size 477408072
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0455d53b1b0c7edcbcf5b7a17816685f33f49d39da811eb1c2efa3dc35fb85c
3
+ size 479350955
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:345d3af616c67a79077f0f636994ea71319caae37d590d07bfc3625fc7dd5ba1
3
- size 476285658
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb80680a2cb37392ad0a1ccb36e291a8ca14fff50848a1fe943110d042c91b95
3
+ size 478228998
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4b1612f3b65f5565745241e15fa8fca805ed339dce89c3d35b5c82fc774f19b8
3
- size 1698575820
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:450dff530ac7190bac7689ecf04f1b5ba3cd484dee7ce073ca2a46009a4fdb2b
3
+ size 1698575515
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f05b0ca08d0e219c48781697ae40aa5663f55016cecbc2c17abfd34c7b62a425
3
- size 425822799
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce085b88aaf7e1bb564a8c8580ff60129e2191ae85404253060457b5c57f0ee3
3
+ size 427520153
onnx/encoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:95c00dc87260918560692d0ab24eaebb4c2f3cd8588a832e29ba7186f86cce68
3
  size 1365283172
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c58209b472cdf3e7fea925f109fe7ee67d12d99368ccd8d18a18c75cce6a9e51
3
  size 1365283172
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:26af9fa17778b65326f0e79746b0cc03b9fc8e3812a1c843b9135a5cf57300b5
3
- size 342049899
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c913a4d58ce74bb978125292667050e426692e95bec67a70388569b3cad82034
3
+ size 343340979
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
+ "Add",
8
+ "Less",
9
+ "ConstantOfShape",
10
+ "Mul",
11
+ "Pow",
12
+ "Gather",
13
+ "Shape",
14
+ "ReduceMean",
15
+ "MatMul",
16
+ "Range",
17
+ "Transpose",
18
+ "Neg",
19
+ "Expand",
20
+ "Sqrt",
21
+ "Min",
22
+ "Unsqueeze",
23
+ "Where",
24
+ "Div",
25
+ "Tile",
26
+ "Log",
27
+ "Concat",
28
+ "Cast",
29
+ "LessOrEqual",
30
+ "Reshape",
31
+ "Sub",
32
+ "Softmax",
33
+ "Tanh",
34
+ "Constant"
35
+ ],
36
+ "weight_type": "QInt8"
37
+ },
38
+ "encoder_model": {
39
+ "op_types": [
40
+ "Add",
41
+ "Abs",
42
+ "Less",
43
+ "ConstantOfShape",
44
+ "Mul",
45
+ "Pow",
46
+ "Gather",
47
+ "Shape",
48
+ "ReduceMean",
49
+ "MatMul",
50
+ "Range",
51
+ "Transpose",
52
+ "Sqrt",
53
+ "Min",
54
+ "Unsqueeze",
55
+ "Where",
56
+ "Div",
57
+ "Greater",
58
+ "Log",
59
+ "Concat",
60
+ "Cast",
61
+ "Reshape",
62
+ "Sub",
63
+ "Softmax",
64
+ "Tanh",
65
+ "Constant"
66
+ ],
67
+ "weight_type": "QInt8"
68
+ },
69
+ "decoder_model_merged": {
70
+ "op_types": [
71
+ "If",
72
+ "Add",
73
+ "Less",
74
+ "ConstantOfShape",
75
+ "Mul",
76
+ "Slice",
77
+ "Pow",
78
+ "Gather",
79
+ "Shape",
80
+ "ReduceMean",
81
+ "MatMul",
82
+ "Range",
83
+ "Transpose",
84
+ "Neg",
85
+ "Expand",
86
+ "Sqrt",
87
+ "Min",
88
+ "Unsqueeze",
89
+ "Where",
90
+ "Div",
91
+ "Tile",
92
+ "Log",
93
+ "Concat",
94
+ "Cast",
95
+ "LessOrEqual",
96
+ "Reshape",
97
+ "Sub",
98
+ "Softmax",
99
+ "Tanh",
100
+ "Constant"
101
+ ],
102
+ "weight_type": "QInt8"
103
+ },
104
+ "decoder_with_past_model": {
105
+ "op_types": [
106
+ "Add",
107
+ "Less",
108
+ "ConstantOfShape",
109
+ "Mul",
110
+ "Slice",
111
+ "Pow",
112
+ "Gather",
113
+ "Shape",
114
+ "ReduceMean",
115
+ "MatMul",
116
+ "Range",
117
+ "Transpose",
118
+ "Neg",
119
+ "Expand",
120
+ "Sqrt",
121
+ "Min",
122
+ "Unsqueeze",
123
+ "Where",
124
+ "Div",
125
+ "Tile",
126
+ "Log",
127
+ "Concat",
128
+ "Cast",
129
+ "LessOrEqual",
130
+ "Reshape",
131
+ "Sub",
132
+ "Softmax",
133
+ "Tanh",
134
+ "Constant"
135
+ ],
136
+ "weight_type": "QInt8"
137
+ }
138
+ }
139
+ }