tarekziade commited on
Commit
5bbea1f
1 Parent(s): b7cc1ca

Upload 18 files

Browse files
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b562d5631d23b71d9417071e8041f3eb80e84122a9bbb66dfdec8d46d91594d4
3
  size 385864797
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5aedbc7bb81581bdd4b90f65db39dc22b3dcdea3923d67542449aa24bd46eaf7
3
  size 385864797
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:129100f907e7295525d4529083bfb5bdcc206eedce9f9d4a6223131ea8cb8e87
3
  size 387342586
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c398f8e6fd8bf8b03105a7ada1541f488c07870229210a128205c6492b7c38ed
3
  size 387342586
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d636dce4f19d08059cf075c25b75f94f71e5b28081d3be8c3823c07f40c59604
3
  size 99759579
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0c66eda97aa444b17357dfba1c9827abc7824d6fd905dd28f097f74ddef02943
3
  size 99759579
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:87de04f08ef3a7cec2f3cd71d4de440864ab0a7f4f32ad0764fdb8d26387fe94
3
  size 98065763
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cda4a51db36003185beb5557cc379d0ad773d4efa9329e39793828b74d420f4
3
  size 98065763
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d3e69a840d8b693b738d31f9bab09434311b1eb022f4c6a89f7c31ae4ac76428
3
  size 385864377
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ab017c9a31793692bb56c0b93be42cb0090d97f6efc9cbd25f092a93451f45e
3
  size 385864377
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5da506c3668083b54468a4be3d0581c2749abc28c68d4f6f816d7f03548ea977
3
  size 98063170
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7885df45251f50971c9a8fdf11c79011342988bcabdb0649470751076226b8ae
3
  size 98063170
onnx/encoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e06aee9b192045a32c40f0bac2e1304e153f7b7521286a65a478893a582b4087
3
  size 343440632
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e48effb1e61fcecf4d1587fb15894d15510f366971ff22224719048168a70707
3
  size 343440632
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:90563de4250ab0e12730242476b884529a1a825059c36212bd4ae1dac2e87c56
3
  size 87038173
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04622bdece4419687c8a07f4df0e8c1bea1db354fb6decde97bc0bf39f26bab2
3
  size 87038173
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3c82a8f09d299196b2a3237374d874d008567e0d9d9044c48b24965138d8394e
3
  size 730052378
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea90951d68491925423f56366196ab4d2a7cbafecc88d58623b8f74ae2118872
3
  size 730052378
quantize_config.json CHANGED
@@ -4,120 +4,120 @@
4
  "per_model_config": {
5
  "decoder_model": {
6
  "op_types": [
7
- "Add",
8
- "Transpose",
9
- "Squeeze",
10
- "Range",
11
- "Slice",
12
- "ConstantOfShape",
13
  "MatMul",
14
- "Where",
15
- "ReduceMean",
 
 
 
 
16
  "Div",
17
- "Unsqueeze",
18
  "Split",
 
 
 
 
 
19
  "Cast",
 
 
 
 
20
  "Gemm",
21
  "Tanh",
22
- "Pow",
23
- "Mul",
24
- "Gather",
25
- "Reshape",
26
- "Softmax",
27
- "Sqrt",
28
- "Shape",
29
- "Constant",
30
- "Sub",
31
- "Concat"
32
  ],
33
  "weight_type": "QInt8"
34
  },
35
  "decoder_model_merged": {
36
  "op_types": [
37
- "Add",
38
- "Transpose",
39
- "Squeeze",
40
- "Range",
41
- "Slice",
42
- "ConstantOfShape",
43
  "MatMul",
44
- "Where",
45
- "ReduceMean",
 
 
 
 
 
46
  "Div",
47
- "Unsqueeze",
48
  "Split",
 
 
 
 
 
49
  "Cast",
 
 
 
 
50
  "Gemm",
51
  "Tanh",
52
- "Pow",
53
- "Mul",
54
- "Gather",
55
- "Reshape",
56
- "Softmax",
57
- "Sqrt",
58
- "Shape",
59
- "Constant",
60
- "Sub",
61
- "If",
62
- "Concat"
63
  ],
64
  "weight_type": "QInt8"
65
  },
66
  "decoder_with_past_model": {
67
  "op_types": [
68
- "Add",
69
- "Transpose",
70
- "Squeeze",
71
- "Range",
72
- "Slice",
73
- "ConstantOfShape",
74
  "MatMul",
75
- "Where",
76
- "ReduceMean",
 
 
 
 
77
  "Div",
78
- "Unsqueeze",
79
  "Split",
 
 
 
 
 
80
  "Cast",
 
 
 
 
81
  "Gemm",
82
  "Tanh",
83
- "Pow",
84
- "Mul",
85
- "Gather",
86
- "Reshape",
87
- "Softmax",
88
- "Sqrt",
89
- "Shape",
90
- "Constant",
91
- "Sub",
92
- "Concat"
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
  "encoder_model": {
97
  "op_types": [
98
- "Add",
99
- "Transpose",
100
- "Equal",
101
- "Slice",
102
- "Expand",
103
- "ConstantOfShape",
104
  "MatMul",
105
- "Where",
106
- "ReduceMean",
107
- "Div",
108
- "Unsqueeze",
109
- "Conv",
110
- "Pow",
111
  "Mul",
112
- "Gather",
113
- "Reshape",
114
- "Softmax",
115
- "Sqrt",
116
- "Shape",
117
  "Constant",
 
 
 
 
 
118
  "Sub",
 
 
 
119
  "Erf",
120
- "Concat"
 
 
 
 
 
 
 
121
  ],
122
  "weight_type": "QUInt8"
123
  }
 
4
  "per_model_config": {
5
  "decoder_model": {
6
  "op_types": [
7
+ "Sqrt",
 
 
 
 
 
8
  "MatMul",
9
+ "Concat",
10
+ "Mul",
11
+ "Constant",
12
+ "Shape",
13
+ "Range",
14
+ "Pow",
15
  "Div",
 
16
  "Split",
17
+ "Sub",
18
+ "Where",
19
+ "Reshape",
20
+ "Gather",
21
+ "Unsqueeze",
22
  "Cast",
23
+ "ReduceMean",
24
+ "Softmax",
25
+ "ConstantOfShape",
26
+ "Slice",
27
  "Gemm",
28
  "Tanh",
29
+ "Transpose",
30
+ "Add",
31
+ "Squeeze"
 
 
 
 
 
 
 
32
  ],
33
  "weight_type": "QInt8"
34
  },
35
  "decoder_model_merged": {
36
  "op_types": [
37
+ "Sqrt",
 
 
 
 
 
38
  "MatMul",
39
+ "Concat",
40
+ "Mul",
41
+ "If",
42
+ "Constant",
43
+ "Shape",
44
+ "Range",
45
+ "Pow",
46
  "Div",
 
47
  "Split",
48
+ "Sub",
49
+ "Where",
50
+ "Reshape",
51
+ "Gather",
52
+ "Unsqueeze",
53
  "Cast",
54
+ "ReduceMean",
55
+ "Softmax",
56
+ "ConstantOfShape",
57
+ "Slice",
58
  "Gemm",
59
  "Tanh",
60
+ "Transpose",
61
+ "Add",
62
+ "Squeeze"
 
 
 
 
 
 
 
 
63
  ],
64
  "weight_type": "QInt8"
65
  },
66
  "decoder_with_past_model": {
67
  "op_types": [
68
+ "Sqrt",
 
 
 
 
 
69
  "MatMul",
70
+ "Concat",
71
+ "Mul",
72
+ "Constant",
73
+ "Shape",
74
+ "Range",
75
+ "Pow",
76
  "Div",
 
77
  "Split",
78
+ "Sub",
79
+ "Where",
80
+ "Reshape",
81
+ "Gather",
82
+ "Unsqueeze",
83
  "Cast",
84
+ "ReduceMean",
85
+ "Softmax",
86
+ "ConstantOfShape",
87
+ "Slice",
88
  "Gemm",
89
  "Tanh",
90
+ "Transpose",
91
+ "Squeeze",
92
+ "Add"
 
 
 
 
 
 
 
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
  "encoder_model": {
97
  "op_types": [
98
+ "Sqrt",
 
 
 
 
 
99
  "MatMul",
100
+ "Concat",
 
 
 
 
 
101
  "Mul",
 
 
 
 
 
102
  "Constant",
103
+ "Expand",
104
+ "Shape",
105
+ "Pow",
106
+ "Div",
107
+ "Conv",
108
  "Sub",
109
+ "Where",
110
+ "Reshape",
111
+ "Gather",
112
  "Erf",
113
+ "Unsqueeze",
114
+ "Equal",
115
+ "ReduceMean",
116
+ "Softmax",
117
+ "ConstantOfShape",
118
+ "Slice",
119
+ "Transpose",
120
+ "Add"
121
  ],
122
  "weight_type": "QUInt8"
123
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c6124d14554c856f07f64ebbca472baff63ce12555bd850dbac15bdf294e64c5
3
  size 4728
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6df8b3132ea43f82797da62dc92bab7492b597315c59b9eeae0937e97904f9e0
3
  size 4728