Xenova HF staff commited on
Commit
c2d50b8
1 Parent(s): f994808

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -39,7 +39,7 @@
39
  "scale_embedding": true,
40
  "share_encoder_decoder_embeddings": true,
41
  "static_position_embeddings": true,
42
- "transformers_version": "4.32.0.dev0",
43
  "use_cache": true,
44
  "vocab_size": 57060
45
  }
 
39
  "scale_embedding": true,
40
  "share_encoder_decoder_embeddings": true,
41
  "static_position_embeddings": true,
42
+ "transformers_version": "4.34.0.dev0",
43
  "use_cache": true,
44
  "vocab_size": 57060
45
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  57059
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 4,
14
  "pad_token_id": 57059,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  57059
 
11
  "max_length": 512,
12
  "num_beams": 4,
13
  "pad_token_id": 57059,
14
+ "renormalize_logits": true,
15
+ "transformers_version": "4.34.0.dev0"
16
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fd8fca7108f0cfe2fd71769ed1b96d1dd52fc1e1d6f85ce72a9fbc8fc2b334d8
3
- size 875253264
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3081e52ccc96463e56be90bfdeb1f74cc5ee7edd2b66217abd3009b702ed53a9
3
+ size 641535562
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:efaa9fd6dfd5facddeb9c6e36df76c28dc982c2ebc20dd49e193aad41ac0c463
3
- size 875487492
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b164bf82dfa332b84cb79ecc71859ee1e34815aab2bfb80aa1b8fe80b0f32d6
3
+ size 641769875
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:76597856c34d278e624f8fcccfb7291807a132815b862d0cd9a84f4bb9879460
3
- size 220760044
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0043b443a8a85427612d3e30ad5228ee6e7c0da4d99d3b831d6f0e6f5dceee6b
3
+ size 162043892
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:edb8d149e96e1d622ee98da1a994affc926b16fa4dc7684267a963a3053947e1
3
- size 220390278
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:49d0bce58fdf9c649cc6c3d0aeb7e2f76afd9622bcb3985825903fe2b56d3fc0
3
+ size 161674780
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5ff5783cadc597de352690762265a1a92969627029f4ba41fb6489ef67186121
3
- size 824829419
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a40e261c0d98bede7f5b9dae6edb5e688798800b044f6f0595e9bf74e1e2673
3
+ size 591111717
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1949f1e57d22ffe30058c8b7a3b4f2623d9d59c1e2066cfa4af50dd87b4b5a33
3
- size 207626334
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fa9c4b2a505d338283fe2af9591d6c9664a53650fe72b244c52dd2a9703334b
3
+ size 148910836
quantize_config.json CHANGED
@@ -4,120 +4,120 @@
4
  "per_model_config": {
5
  "encoder_model": {
6
  "op_types": [
7
- "Transpose",
8
- "Shape",
9
  "Where",
10
- "Div",
11
- "ReduceMean",
 
 
 
12
  "Relu",
13
- "Softmax",
 
 
14
  "Sqrt",
15
- "ConstantOfShape",
16
- "Mul",
 
 
17
  "Cast",
18
  "MatMul",
19
- "Range",
20
- "Concat",
21
- "Reshape",
22
- "Expand",
23
- "Gather",
24
- "Unsqueeze",
25
  "Add",
26
- "Constant",
27
- "Sub",
28
- "Pow",
29
- "Equal"
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
  "decoder_with_past_model": {
34
  "op_types": [
35
- "Transpose",
36
- "Shape",
37
  "Where",
38
- "Div",
39
- "ReduceMean",
 
 
 
40
  "Relu",
41
- "Softmax",
 
 
42
  "Sqrt",
43
- "ConstantOfShape",
44
- "Mul",
 
 
45
  "Cast",
46
  "MatMul",
47
- "Range",
48
- "Concat",
49
- "Reshape",
50
- "Expand",
51
- "Gather",
52
- "Unsqueeze",
53
  "Add",
54
- "Constant",
55
- "Sub",
56
- "Pow",
57
- "Equal"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
63
- "Transpose",
64
- "Shape",
65
  "Where",
66
- "Div",
67
- "ReduceMean",
 
 
 
 
68
  "Relu",
69
- "Softmax",
 
 
70
  "Sqrt",
71
- "ConstantOfShape",
72
- "Mul",
73
- "Cast",
74
- "MatMul",
75
  "Range",
76
- "Concat",
77
- "Reshape",
78
  "Less",
79
- "Slice",
 
 
80
  "Expand",
81
- "Gather",
82
- "Unsqueeze",
83
  "Squeeze",
84
- "Add",
85
- "Constant",
86
- "Sub",
87
- "Pow",
88
- "Equal"
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
94
- "Transpose",
95
- "Shape",
96
  "Where",
97
- "Div",
98
- "ReduceMean",
 
 
 
99
  "Relu",
100
- "Softmax",
 
 
101
  "Sqrt",
102
- "ConstantOfShape",
103
- "Mul",
104
- "Cast",
105
- "MatMul",
106
  "Range",
107
- "If",
108
- "Concat",
109
- "Reshape",
110
  "Less",
111
- "Slice",
 
 
112
  "Expand",
113
- "Gather",
114
- "Unsqueeze",
115
  "Squeeze",
116
- "Add",
117
- "Constant",
118
- "Sub",
119
- "Pow",
120
- "Equal"
121
  ],
122
  "weight_type": "QInt8"
123
  }
 
4
  "per_model_config": {
5
  "encoder_model": {
6
  "op_types": [
7
+ "Sub",
 
8
  "Where",
9
+ "Equal",
10
+ "Shape",
11
+ "Reshape",
12
+ "ConstantOfShape",
13
+ "Pow",
14
  "Relu",
15
+ "Div",
16
+ "Gather",
17
+ "Constant",
18
  "Sqrt",
19
+ "Softmax",
20
+ "Unsqueeze",
21
+ "Transpose",
22
+ "Range",
23
  "Cast",
24
  "MatMul",
 
 
 
 
 
 
25
  "Add",
26
+ "Mul",
27
+ "Expand",
28
+ "Concat",
29
+ "ReduceMean"
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
  "decoder_with_past_model": {
34
  "op_types": [
35
+ "Sub",
 
36
  "Where",
37
+ "Equal",
38
+ "Shape",
39
+ "Reshape",
40
+ "ConstantOfShape",
41
+ "Pow",
42
  "Relu",
43
+ "Div",
44
+ "Gather",
45
+ "Constant",
46
  "Sqrt",
47
+ "Softmax",
48
+ "Unsqueeze",
49
+ "Transpose",
50
+ "Range",
51
  "Cast",
52
  "MatMul",
 
 
 
 
 
 
53
  "Add",
54
+ "Mul",
55
+ "Expand",
56
+ "Concat",
57
+ "ReduceMean"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
+ "decoder_model_merged": {
62
  "op_types": [
63
+ "Sub",
 
64
  "Where",
65
+ "Equal",
66
+ "If",
67
+ "Shape",
68
+ "Reshape",
69
+ "ConstantOfShape",
70
+ "Pow",
71
  "Relu",
72
+ "Div",
73
+ "Gather",
74
+ "Constant",
75
  "Sqrt",
76
+ "Softmax",
77
+ "Unsqueeze",
78
+ "Transpose",
79
+ "Slice",
80
  "Range",
81
+ "Cast",
 
82
  "Less",
83
+ "Add",
84
+ "MatMul",
85
+ "Mul",
86
  "Expand",
 
 
87
  "Squeeze",
88
+ "Concat",
89
+ "ReduceMean"
 
 
 
90
  ],
91
  "weight_type": "QInt8"
92
  },
93
+ "decoder_model": {
94
  "op_types": [
95
+ "Sub",
 
96
  "Where",
97
+ "Equal",
98
+ "Shape",
99
+ "Reshape",
100
+ "ConstantOfShape",
101
+ "Pow",
102
  "Relu",
103
+ "Div",
104
+ "Gather",
105
+ "Constant",
106
  "Sqrt",
107
+ "Softmax",
108
+ "Unsqueeze",
109
+ "Transpose",
110
+ "Slice",
111
  "Range",
112
+ "Cast",
 
 
113
  "Less",
114
+ "MatMul",
115
+ "Add",
116
+ "Mul",
117
  "Expand",
 
 
118
  "Squeeze",
119
+ "Concat",
120
+ "ReduceMean"
 
 
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff