Xenova HF staff commited on
Commit
2afa3df
1 Parent(s): f7cdab8

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -139,8 +139,7 @@
139
  50360,
140
  50361
141
  ],
142
- "torch_dtype": "float32",
143
- "transformers_version": "4.29.2",
144
  "use_cache": true,
145
  "use_weighted_layer_sum": false,
146
  "vocab_size": 51864
 
139
  50360,
140
  50361
141
  ],
142
+ "transformers_version": "4.30.2",
 
143
  "use_cache": true,
144
  "use_weighted_layer_sum": false,
145
  "vocab_size": 51864
generation_config.json CHANGED
@@ -110,5 +110,5 @@
110
  50360,
111
  50361
112
  ],
113
- "transformers_version": "4.29.2"
114
  }
 
110
  50360,
111
  50361
112
  ],
113
+ "transformers_version": "4.30.2"
114
  }
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9e85f9a29a0b0e1dfdb2211ac09ff1bda4faa5808b5a9855a843982367ac58c2
3
- size 198194454
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:441da7a85e2cac5508fad6795cf115333b8566cffc886326788e0a64366411a5
3
+ size 198194466
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8b79745272590fbc4e841188d1ce522efbca6d593a7dbbc35a56b87908b211ee
3
- size 50942904
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29e4ab70caaedbfd18a4cc4fab3d0708b9eb7c888f54f31f11949a6a040fcf36
3
+ size 50583569
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7e845d7c9dc4317bfe89fd01e6f2cd14fa0b045cd099f229ee067ad9f62a0ec7
3
- size 50707084
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:699a1478cd7b1126a8f345281c1fb29fa72a18b8f45e4b31c6499653e3dee1d2
3
+ size 50347713
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:11085368e8f75c52e9ed52d75f1bdf149aa9f0dcb29f9f2a3795975fb755f08a
3
- size 49457772
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1182513a4c3ccd16fbbabb2a760f86eac2e827d9f8ce4c78c33e3935af86526c
3
+ size 49113801
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8149b739060eaa30b6211a5e107e33e422f5c4d914229b94357f31bac4ed68f0
3
- size 10182475
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:66bb0bf174a3a3bcfdaccea161c811d0d90a25c927021ab9ffbf639b902225f1
3
+ size 10113251
quant_config.json ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": false,
3
+ "reduce_range": false,
4
+ "per_model_config": {
5
+ "encoder_model": {
6
+ "op_types": [
7
+ "Transpose",
8
+ "ReduceMean",
9
+ "Sub",
10
+ "Gather",
11
+ "Conv",
12
+ "Shape",
13
+ "Pow",
14
+ "MatMul",
15
+ "Div",
16
+ "Erf",
17
+ "Softmax",
18
+ "Reshape",
19
+ "Concat",
20
+ "Add",
21
+ "Unsqueeze",
22
+ "Constant",
23
+ "Mul",
24
+ "Sqrt"
25
+ ],
26
+ "weight_type": "QUInt8"
27
+ },
28
+ "decoder_model": {
29
+ "op_types": [
30
+ "ReduceMean",
31
+ "Sub",
32
+ "Pow",
33
+ "Cast",
34
+ "MatMul",
35
+ "Softmax",
36
+ "Where",
37
+ "Range",
38
+ "Equal",
39
+ "Less",
40
+ "Div",
41
+ "Erf",
42
+ "Expand",
43
+ "Concat",
44
+ "Mul",
45
+ "Sqrt",
46
+ "Transpose",
47
+ "Gather",
48
+ "Shape",
49
+ "ConstantOfShape",
50
+ "Slice",
51
+ "Squeeze",
52
+ "Reshape",
53
+ "Add",
54
+ "Unsqueeze",
55
+ "Constant"
56
+ ],
57
+ "weight_type": "QInt8"
58
+ },
59
+ "decoder_model_merged": {
60
+ "op_types": [
61
+ "If"
62
+ ],
63
+ "weight_type": "QInt8"
64
+ },
65
+ "decoder_with_past_model": {
66
+ "op_types": [
67
+ "ReduceMean",
68
+ "Gather",
69
+ "Sub",
70
+ "Transpose",
71
+ "Shape",
72
+ "Slice",
73
+ "Pow",
74
+ "MatMul",
75
+ "Div",
76
+ "Softmax",
77
+ "Erf",
78
+ "Reshape",
79
+ "Concat",
80
+ "Add",
81
+ "Unsqueeze",
82
+ "Constant",
83
+ "Mul",
84
+ "Sqrt"
85
+ ],
86
+ "weight_type": "QInt8"
87
+ }
88
+ }
89
+ }