Xenova HF staff commited on
Commit
d3754ac
1 Parent(s): 46c30e7

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 sha256:227d7f58e4f7d838ccb0bfda118414ac33bfaaf8fbdb72679b42b0c5d4d125a8
3
- size 774224014
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9127320059559eb7c1d39b10c573c4653a3f981741ccc3e8fc3b1ef911b3f433
3
+ size 614895553
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9264951f2068dd6681aab91b25492475433c93272e5e49bc7adcad3a6ead31ad
3
- size 774732105
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:54ed870f406c72bd6785bd79235a51936351424332b56ddd178471915da61b9c
3
+ size 615405212
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ff3ea98a0f2b257cc378f4da2d230fc4f9353ba6761f2d7c0907a5f3fd3c4d40
3
- size 196611875
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5b54903e443ce173b1465b420442029e32a19dc474e4044eddd6acaa0005ea5
3
+ size 156780950
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:473e5b1fadaf85de08c7eab8739425f9dc291b74a62db6295f68fb1f1c68ea26
3
- size 195823308
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65f7eec90fb9dc716236c491fed03d7b169c8f2c4c67d99df1cfe0a1bda3246e
3
+ size 155991279
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7da31758b812c7896df6ebb1e4b198b25acdd855661b7f7304572566bca7ae71
3
- size 717493595
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4be9d10f9ba006f3e8230b2d9b87a114b14ab679be346a6a43b5b3ebe2fe8ab0
3
+ size 558165134
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:690e58d58467b829e0d33c684913f550945fadf88f0a6947f545f9919262b09d
3
- size 181511486
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfb9423243362829d2252050ac133775cfc6f1b8229d074db1548f7547ff4a02
3
+ size 141679457
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
- "ReduceMean",
8
- "Transpose",
9
- "Pow",
10
- "Softmax",
11
- "MatMul",
12
  "Erf",
13
- "Constant",
14
- "Reshape",
15
  "Cast",
 
16
  "Squeeze",
 
 
 
 
 
 
17
  "Div",
 
 
 
 
18
  "Expand",
19
- "Sub",
 
20
  "Slice",
21
- "Add",
22
  "ConstantOfShape",
23
- "Equal",
24
- "Unsqueeze",
25
- "Range",
26
- "Gather",
27
- "If",
28
- "Sqrt",
29
- "Mul",
30
  "Where",
31
- "Concat",
32
- "Less",
33
- "Shape"
 
34
  ],
35
  "weight_type": "QInt8"
36
  },
37
- "decoder_model": {
38
  "op_types": [
39
- "ReduceMean",
40
- "Transpose",
41
- "Pow",
42
- "Softmax",
43
- "MatMul",
44
  "Erf",
45
- "Constant",
46
- "Reshape",
47
  "Cast",
 
48
  "Squeeze",
 
 
 
 
 
 
49
  "Div",
 
 
 
 
50
  "Expand",
51
- "Sub",
 
52
  "Slice",
53
- "Add",
54
  "ConstantOfShape",
55
- "Equal",
56
- "Unsqueeze",
57
- "Range",
58
- "Gather",
59
- "Sqrt",
60
- "Mul",
61
  "Where",
 
 
 
62
  "Concat",
63
- "Less",
64
- "Shape"
65
  ],
66
  "weight_type": "QInt8"
67
  },
68
  "decoder_with_past_model": {
69
  "op_types": [
70
- "ReduceMean",
71
- "Sqrt",
72
- "Transpose",
73
- "Div",
74
- "Unsqueeze",
75
- "Pow",
76
  "Mul",
77
- "MatMul",
78
- "Constant",
79
- "Softmax",
80
- "Concat",
81
  "Erf",
82
  "Gather",
 
83
  "Reshape",
 
 
84
  "Sub",
 
 
 
 
 
85
  "Slice",
86
- "Shape",
87
- "Add"
 
 
88
  ],
89
  "weight_type": "QInt8"
90
  },
91
  "encoder_model": {
92
  "op_types": [
93
- "ReduceMean",
94
- "Transpose",
95
- "Div",
96
- "Pow",
97
- "Unsqueeze",
98
  "Erf",
99
  "Mul",
100
- "Conv",
101
- "Constant",
102
- "MatMul",
103
- "Concat",
104
- "Softmax",
105
  "Shape",
106
  "Gather",
107
  "Reshape",
 
 
108
  "Sub",
 
 
 
 
 
109
  "Sqrt",
110
- "Add"
 
 
 
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
+ "Equal",
 
9
  "Cast",
10
+ "Sqrt",
11
  "Squeeze",
12
+ "ReduceMean",
13
+ "Mul",
14
+ "Gather",
15
+ "Sub",
16
+ "Pow",
17
+ "Transpose",
18
  "Div",
19
+ "Less",
20
+ "Shape",
21
+ "Reshape",
22
+ "Softmax",
23
  "Expand",
24
+ "MatMul",
25
+ "Range",
26
  "Slice",
 
27
  "ConstantOfShape",
 
 
 
 
 
 
 
28
  "Where",
29
+ "Constant",
30
+ "Add",
31
+ "Unsqueeze",
32
+ "Concat"
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "decoder_model_merged": {
37
  "op_types": [
 
 
 
 
 
38
  "Erf",
39
+ "Equal",
 
40
  "Cast",
41
+ "Sqrt",
42
  "Squeeze",
43
+ "ReduceMean",
44
+ "Mul",
45
+ "Gather",
46
+ "Sub",
47
+ "Pow",
48
+ "Transpose",
49
  "Div",
50
+ "Less",
51
+ "Shape",
52
+ "Reshape",
53
+ "Softmax",
54
  "Expand",
55
+ "MatMul",
56
+ "Range",
57
  "Slice",
 
58
  "ConstantOfShape",
 
 
 
 
 
 
59
  "Where",
60
+ "Constant",
61
+ "Add",
62
+ "Unsqueeze",
63
  "Concat",
64
+ "If"
 
65
  ],
66
  "weight_type": "QInt8"
67
  },
68
  "decoder_with_past_model": {
69
  "op_types": [
 
 
 
 
 
 
70
  "Mul",
 
 
 
 
71
  "Erf",
72
  "Gather",
73
+ "Shape",
74
  "Reshape",
75
+ "Concat",
76
+ "ReduceMean",
77
  "Sub",
78
+ "Softmax",
79
+ "MatMul",
80
+ "Pow",
81
+ "Sqrt",
82
+ "Transpose",
83
  "Slice",
84
+ "Div",
85
+ "Constant",
86
+ "Add",
87
+ "Unsqueeze"
88
  ],
89
  "weight_type": "QInt8"
90
  },
91
  "encoder_model": {
92
  "op_types": [
 
 
 
 
 
93
  "Erf",
94
  "Mul",
 
 
 
 
 
95
  "Shape",
96
  "Gather",
97
  "Reshape",
98
+ "ReduceMean",
99
+ "Concat",
100
  "Sub",
101
+ "Softmax",
102
+ "Conv",
103
+ "MatMul",
104
+ "Pow",
105
+ "Transpose",
106
  "Sqrt",
107
+ "Div",
108
+ "Constant",
109
+ "Add",
110
+ "Unsqueeze"
111
  ],
112
  "weight_type": "QUInt8"
113
  }