Xenova HF staff commited on
Commit
868283c
1 Parent(s): 5c11d0e

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:c318453eb4bdb7205e9a33c398de2e9db0e663513279fb065bccf4b00aa973ac
3
- size 708438
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71fb71c5ded414ab8f1cbd7d5746d6f52d8e9083205bb346528ca536ff80c849
3
+ size 708465
onnx/decoder_model.onnx_data CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:046f1802b6053dc36e20f2058b1dca128cd9164cd37d34495962afa1f21ca102
3
- size 3264462848
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42be12f0b0f33f649e591eb8c10942adab6f5a94d541a2edd96da1b32e29f62e
3
+ size 2236858368
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7b5440b416feb4f24695692b176225f518deafed3c017cea8ba6d5fe0151cf81
3
- size 1399393
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed3c777abc0adc81cdd5558121b13e40a728edbd75f71da8492914dd99edd153
3
+ size 1399544
onnx/decoder_model_merged.onnx_data CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:046f1802b6053dc36e20f2058b1dca128cd9164cd37d34495962afa1f21ca102
3
- size 3264462848
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42be12f0b0f33f649e591eb8c10942adab6f5a94d541a2edd96da1b32e29f62e
3
+ size 2236858368
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a93ddbe77bb8359c687418bf7592b46f5bd3e5ea2fe8dbf62ea682c73b8cc3ea
3
- size 819144015
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f2cc04e9eff2b1f9045ce3211ed573375c56922ac42a6815f66c31d2143b260
3
+ size 562242270
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e3f054fdc6afd82cbbf5f23b6f32770fff798d7620e2457c838fb5b5a0facec3
3
- size 818116171
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:467e80d6596d6164468e5f48a8adcfde2fd2965863bac47318421ccd751197c3
3
+ size 561214692
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:78714ec76dae1951e26d8823252c78ab621ef2d449fb59d19bcd95d9c7ddd498
3
- size 714961
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18e90698757af8933ef57252d2a16b3bb48752b8d7d0fa6cfa01c14b92b52d75
3
+ size 714978
onnx/decoder_with_past_model.onnx_data CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:046f1802b6053dc36e20f2058b1dca128cd9164cd37d34495962afa1f21ca102
3
- size 3264462848
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42be12f0b0f33f649e591eb8c10942adab6f5a94d541a2edd96da1b32e29f62e
3
+ size 2236858368
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7ffe8bf5750be07cc010c1d018e16e45d13fe8f2ecdde7ae904693fe6a31055a
3
- size 818118299
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94e50cc2acb0141cea2ec728aca9f880f4c45a6260279b0afc590fd711e82e7c
3
+ size 561216820
quantize_config.json CHANGED
@@ -4,99 +4,99 @@
4
  "per_model_config": {
5
  "decoder_model": {
6
  "op_types": [
7
- "ScatterND",
8
- "ReduceMean",
9
- "Transpose",
10
- "Expand",
11
  "CumSum",
12
- "Softmax",
13
- "Gather",
14
  "Reshape",
15
- "Tanh",
16
  "Unsqueeze",
17
- "Cast",
18
  "Pow",
19
- "Equal",
 
 
 
 
 
 
20
  "Concat",
21
- "Constant",
22
- "Slice",
23
- "Sub",
24
  "Range",
 
25
  "Sqrt",
26
- "Not",
27
- "Mul",
28
- "Div",
29
- "Less",
30
- "Or",
31
- "MatMul",
32
- "Shape",
33
- "Add",
34
  "Where",
35
- "ConstantOfShape"
 
 
 
 
 
 
36
  ],
37
  "weight_type": "QInt8"
38
  },
39
  "decoder_model_merged": {
40
  "op_types": [
41
- "ScatterND",
42
- "ReduceMean",
43
- "Transpose",
44
- "Expand",
45
  "CumSum",
46
- "Softmax",
47
- "Gather",
48
  "Reshape",
49
- "Tanh",
50
  "Unsqueeze",
51
- "Cast",
52
  "Pow",
53
- "Equal",
 
 
 
 
 
 
54
  "Concat",
55
- "Constant",
56
- "Slice",
57
- "Sub",
58
  "Range",
 
59
  "Sqrt",
60
- "Not",
61
- "Mul",
 
 
 
62
  "If",
63
- "Div",
 
 
64
  "Less",
65
- "Or",
66
- "MatMul",
67
- "Shape",
68
- "Add",
69
- "Where",
70
- "ConstantOfShape"
71
  ],
72
  "weight_type": "QInt8"
73
  },
74
  "decoder_with_past_model": {
75
  "op_types": [
76
- "ReduceMean",
77
- "Transpose",
78
- "Expand",
79
  "CumSum",
80
- "Softmax",
81
- "Gather",
82
  "Reshape",
83
- "Tanh",
84
  "Unsqueeze",
85
- "Cast",
86
  "Pow",
87
- "Equal",
 
 
 
 
88
  "Concat",
89
- "Constant",
90
- "Sub",
91
  "Sqrt",
92
- "Not",
93
- "Mul",
94
- "Div",
95
- "MatMul",
96
- "Shape",
97
- "Add",
98
  "Where",
99
- "ConstantOfShape"
 
 
 
 
 
100
  ],
101
  "weight_type": "QInt8"
102
  }
 
4
  "per_model_config": {
5
  "decoder_model": {
6
  "op_types": [
7
+ "Add",
 
 
 
8
  "CumSum",
9
+ "Mul",
10
+ "Not",
11
  "Reshape",
12
+ "Div",
13
  "Unsqueeze",
 
14
  "Pow",
15
+ "Or",
16
+ "Shape",
17
+ "Cast",
18
+ "Transpose",
19
+ "MatMul",
20
+ "ScatterND",
21
+ "ReduceMean",
22
  "Concat",
 
 
 
23
  "Range",
24
+ "ConstantOfShape",
25
  "Sqrt",
26
+ "Slice",
27
+ "Constant",
 
 
 
 
 
 
28
  "Where",
29
+ "Expand",
30
+ "Sub",
31
+ "Tanh",
32
+ "Softmax",
33
+ "Equal",
34
+ "Less",
35
+ "Gather"
36
  ],
37
  "weight_type": "QInt8"
38
  },
39
  "decoder_model_merged": {
40
  "op_types": [
41
+ "Add",
 
 
 
42
  "CumSum",
43
+ "Mul",
44
+ "Not",
45
  "Reshape",
46
+ "Div",
47
  "Unsqueeze",
 
48
  "Pow",
49
+ "Or",
50
+ "Shape",
51
+ "Cast",
52
+ "Transpose",
53
+ "MatMul",
54
+ "ScatterND",
55
+ "ReduceMean",
56
  "Concat",
 
 
 
57
  "Range",
58
+ "ConstantOfShape",
59
  "Sqrt",
60
+ "Slice",
61
+ "Constant",
62
+ "Where",
63
+ "Expand",
64
+ "Sub",
65
  "If",
66
+ "Tanh",
67
+ "Softmax",
68
+ "Equal",
69
  "Less",
70
+ "Gather"
 
 
 
 
 
71
  ],
72
  "weight_type": "QInt8"
73
  },
74
  "decoder_with_past_model": {
75
  "op_types": [
76
+ "Add",
 
 
77
  "CumSum",
78
+ "Mul",
79
+ "Not",
80
  "Reshape",
81
+ "Div",
82
  "Unsqueeze",
 
83
  "Pow",
84
+ "Shape",
85
+ "Cast",
86
+ "Transpose",
87
+ "MatMul",
88
+ "ReduceMean",
89
  "Concat",
90
+ "ConstantOfShape",
 
91
  "Sqrt",
92
+ "Constant",
 
 
 
 
 
93
  "Where",
94
+ "Expand",
95
+ "Sub",
96
+ "Tanh",
97
+ "Softmax",
98
+ "Equal",
99
+ "Gather"
100
  ],
101
  "weight_type": "QInt8"
102
  }