Xenova HF staff commited on
Commit
df5325a
1 Parent(s): 237c496

Upload folder using huggingface_hub

Browse files
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f0df994f9b7cf44953eddccdf1084dcf539444a8822678f7a7c9091b46c8a06a
3
- size 2040144906
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d924e4c97da87a8fd8e42b5520f30f426c3dcc0a09729cec954f22b5c9524642
3
+ size 1827707405
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e4d5f56f2ed8b7dbe4b5af5170e561d8d322ca428dd69049dc33b9e564c2e5b9
3
- size 2203713
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efb224b4c05df389b77d0808e5ece453c435fd295343a5982ec87b112cf6250d
3
+ size 1828728265
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b087d7945668de76ab286879dfc314169c2afc52187a84d1f0af0fc9dd7e32aa
3
- size 515768489
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78075f75e4e1bc4a7dc7b21d2b5b902510392481493636e2a26532e6dccdfbd6
3
+ size 462661606
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d82fd8d7d01f812430606d24e4854ecb9408ac81c8b58b1d871fc3c12ce58b7d
3
- size 514188706
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5716d530b10295150db8ae31ee9d154db1a9da0513da7e227e154deea1846f23
3
+ size 461079957
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c2a978fcf70e78f5713dda4738148b0bbb50c316b3424622bc73b43175af4309
3
- size 1838584195
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ee7e218cba90c0b0be92cb714c07eabf53e13f872fe0bf7679d8c68976dc1ed
3
+ size 1626146694
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1da4ca9c388553b37fd364e237123056b97f2fd35a22fa1e64c8324b5118325
3
- size 463527428
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d444eeb3f818e5f455ed5a060ebc10192b7fb0f6d2861fdbfa0d7f882bd523ac
3
+ size 410418679
quantize_config.json CHANGED
@@ -2,112 +2,112 @@
2
  "per_channel": false,
3
  "reduce_range": false,
4
  "per_model_config": {
5
- "decoder_model_merged": {
6
  "op_types": [
7
- "Div",
8
- "Squeeze",
9
- "Pow",
10
  "MatMul",
11
- "Range",
12
- "Equal",
13
- "Softmax",
14
- "Slice",
15
- "If",
16
- "Reshape",
17
  "ReduceMean",
 
18
  "Shape",
19
- "Concat",
20
- "ConstantOfShape",
21
- "Constant",
22
- "Unsqueeze",
23
  "Cast",
24
- "Expand",
25
- "Transpose",
26
- "Gather",
27
- "Less",
28
- "Mul",
29
  "Sub",
 
 
 
30
  "Add",
 
 
 
31
  "Sqrt",
32
- "Erf",
33
- "Where"
 
 
 
 
 
 
 
 
34
  ],
35
  "weight_type": "QInt8"
36
  },
37
- "decoder_model": {
38
  "op_types": [
39
- "Div",
40
- "Squeeze",
41
- "Pow",
42
  "MatMul",
43
- "Range",
44
- "Equal",
45
- "Softmax",
46
- "Slice",
47
- "Reshape",
48
  "ReduceMean",
 
49
  "Shape",
50
- "Concat",
51
- "ConstantOfShape",
52
- "Constant",
53
- "Unsqueeze",
54
  "Cast",
55
- "Expand",
56
- "Transpose",
57
- "Gather",
58
- "Less",
59
- "Mul",
60
  "Sub",
 
 
 
61
  "Add",
 
 
 
62
  "Sqrt",
63
- "Erf",
64
- "Where"
 
 
 
 
 
 
 
 
 
65
  ],
66
  "weight_type": "QInt8"
67
  },
68
  "decoder_with_past_model": {
69
  "op_types": [
 
 
 
70
  "Shape",
 
 
 
 
 
71
  "Transpose",
72
- "Gather",
73
  "Softmax",
74
  "Concat",
75
- "Mul",
 
 
76
  "Div",
77
  "Constant",
78
- "Unsqueeze",
79
- "Slice",
80
- "Pow",
81
- "MatMul",
82
- "Sub",
83
- "Add",
84
- "Sqrt",
85
- "Erf",
86
- "Reshape",
87
- "ReduceMean"
88
  ],
89
  "weight_type": "QInt8"
90
  },
91
  "encoder_model": {
92
  "op_types": [
93
- "Shape",
 
 
 
94
  "Transpose",
95
- "ReduceMean",
96
- "Gather",
 
 
97
  "Softmax",
98
  "Concat",
99
- "Mul",
 
 
100
  "Div",
101
  "Constant",
102
- "Unsqueeze",
103
- "Pow",
104
- "MatMul",
105
- "Sub",
106
- "Add",
107
- "Sqrt",
108
- "Erf",
109
- "Reshape",
110
- "Conv"
111
  ],
112
  "weight_type": "QUInt8"
113
  }
 
2
  "per_channel": false,
3
  "reduce_range": false,
4
  "per_model_config": {
5
+ "decoder_model": {
6
  "op_types": [
7
+ "Erf",
 
 
8
  "MatMul",
9
+ "Transpose",
 
 
 
 
 
10
  "ReduceMean",
11
+ "Div",
12
  "Shape",
 
 
 
 
13
  "Cast",
 
 
 
 
 
14
  "Sub",
15
+ "ConstantOfShape",
16
+ "Range",
17
+ "Where",
18
  "Add",
19
+ "Expand",
20
+ "Pow",
21
+ "Squeeze",
22
  "Sqrt",
23
+ "Gather",
24
+ "Mul",
25
+ "Less",
26
+ "Constant",
27
+ "Unsqueeze",
28
+ "Reshape",
29
+ "Slice",
30
+ "Softmax",
31
+ "Concat",
32
+ "Equal"
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "decoder_model_merged": {
37
  "op_types": [
38
+ "Erf",
 
 
39
  "MatMul",
40
+ "Transpose",
 
 
 
 
41
  "ReduceMean",
42
+ "Div",
43
  "Shape",
 
 
 
 
44
  "Cast",
 
 
 
 
 
45
  "Sub",
46
+ "ConstantOfShape",
47
+ "Range",
48
+ "Where",
49
  "Add",
50
+ "Expand",
51
+ "Pow",
52
+ "Squeeze",
53
  "Sqrt",
54
+ "Gather",
55
+ "Mul",
56
+ "Less",
57
+ "Constant",
58
+ "Unsqueeze",
59
+ "Reshape",
60
+ "Slice",
61
+ "Softmax",
62
+ "Concat",
63
+ "Equal",
64
+ "If"
65
  ],
66
  "weight_type": "QInt8"
67
  },
68
  "decoder_with_past_model": {
69
  "op_types": [
70
+ "Unsqueeze",
71
+ "Erf",
72
+ "MatMul",
73
  "Shape",
74
+ "Reshape",
75
+ "Add",
76
+ "Slice",
77
+ "Sub",
78
+ "Pow",
79
  "Transpose",
 
80
  "Softmax",
81
  "Concat",
82
+ "Sqrt",
83
+ "Gather",
84
+ "ReduceMean",
85
  "Div",
86
  "Constant",
87
+ "Mul"
 
 
 
 
 
 
 
 
 
88
  ],
89
  "weight_type": "QInt8"
90
  },
91
  "encoder_model": {
92
  "op_types": [
93
+ "Unsqueeze",
94
+ "Erf",
95
+ "MatMul",
96
+ "Add",
97
  "Transpose",
98
+ "Sub",
99
+ "Pow",
100
+ "Shape",
101
+ "Reshape",
102
  "Softmax",
103
  "Concat",
104
+ "Sqrt",
105
+ "Gather",
106
+ "ReduceMean",
107
  "Div",
108
  "Constant",
109
+ "Conv",
110
+ "Mul"
 
 
 
 
 
 
 
111
  ],
112
  "weight_type": "QUInt8"
113
  }