Xenova HF staff commited on
Commit
904d8e4
1 Parent(s): e555d1a

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -54,7 +54,7 @@
54
  "scale_embedding": true,
55
  "share_encoder_decoder_embeddings": true,
56
  "static_position_embeddings": true,
57
- "transformers_version": "4.32.0.dev0",
58
  "use_cache": true,
59
  "vocab_size": 58101
60
  }
 
54
  "scale_embedding": true,
55
  "share_encoder_decoder_embeddings": true,
56
  "static_position_embeddings": true,
57
+ "transformers_version": "4.34.0.dev0",
58
  "use_cache": true,
59
  "vocab_size": 58101
60
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  58100
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 4,
14
  "pad_token_id": 58100,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  58100
 
11
  "max_length": 512,
12
  "num_beams": 4,
13
  "pad_token_id": 58100,
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 sha256:426566f247bed99401fbf8260a35f61bf6bfb36ac3aea831919cdf724752f61f
3
- size 340435660
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cc1879f86337fa4278a8c2139eab7a4917c52d076afe6d52d7c9c26dae358ff
3
+ size 221444869
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b7a1d3d99e78835f4ff68d1ce559f190ef5b538eb5055e77690aac34e587ff3e
3
- size 340671142
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31016c44da6f977f6990f03f854be1b3de400a7bbf288f765f2886d445869863
3
+ size 221680436
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a8faafdd7532118635dc143ce4bdbfd2f56384fb35a8c13699c84b7e5abd5595
3
- size 86692033
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4712800584b2d77c2f9e620bf29523e2b20b62f65ade5011ee01780e2b0ea5d3
3
+ size 56652404
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cef6a30562f0c89095772d02a9b5e273fcae200512c9e8c25bdb60eaf00369de
3
- size 86320677
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e44c1c4b50e8f51e49d4d5e54a9af1550dc74763d4023644a38af62611e6efc9
3
+ size 56281702
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4f13568ec0e52fc5ab603602061da9cfcf147912f7936327581ea3d5927b9ac1
3
- size 327785151
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83e23c57592337878ccbb5f7586c5a2d6ae0216f56e11f8f242965fa5f6234a3
3
+ size 208794360
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:413945019dad89cda9f03564880ed6fce2efe8209db95792a025b48c4248d3c6
3
- size 83049213
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e7750192fbfbe0299246cea4f74567a578903fe776e8f64cafb9d02386d6eb0
3
+ size 53010238
quantize_config.json CHANGED
@@ -2,122 +2,122 @@
2
  "per_channel": true,
3
  "reduce_range": true,
4
  "per_model_config": {
5
- "encoder_model": {
6
  "op_types": [
7
- "ReduceMean",
8
- "Reshape",
9
- "Div",
10
- "Add",
11
  "Sub",
12
- "MatMul",
13
- "Constant",
14
  "Mul",
15
- "Range",
16
- "Gather",
17
- "Expand",
18
  "Sigmoid",
19
- "ConstantOfShape",
 
20
  "Equal",
21
- "Sqrt",
22
- "Cast",
 
23
  "Where",
24
- "Unsqueeze",
25
- "Softmax",
26
- "Pow",
27
  "Transpose",
28
- "Concat",
29
- "Shape"
 
 
 
 
 
 
 
 
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
- "decoder_with_past_model": {
34
  "op_types": [
35
- "ReduceMean",
36
- "Reshape",
37
- "Div",
38
  "Sub",
39
- "Add",
40
- "MatMul",
41
- "Constant",
42
  "Mul",
43
- "Range",
44
- "Gather",
45
- "Expand",
46
  "Sigmoid",
47
- "ConstantOfShape",
 
48
  "Equal",
49
- "Sqrt",
50
- "Cast",
51
  "Where",
52
- "Unsqueeze",
53
- "Softmax",
54
- "Pow",
55
  "Transpose",
56
- "Concat",
57
- "Shape"
 
 
 
 
 
 
 
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
63
- "ReduceMean",
64
- "Reshape",
65
- "Div",
66
- "Slice",
67
- "Add",
68
  "Sub",
69
- "MatMul",
70
- "Constant",
71
  "Mul",
72
- "Range",
73
  "Less",
74
- "Gather",
75
- "Expand",
76
  "Sigmoid",
77
- "ConstantOfShape",
 
 
78
  "Equal",
79
- "Sqrt",
80
- "Cast",
 
81
  "Where",
82
- "Unsqueeze",
83
- "Softmax",
84
- "Pow",
85
  "Transpose",
86
- "Concat",
87
- "Shape",
88
- "Squeeze"
 
 
 
 
 
 
 
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
94
- "ReduceMean",
95
- "Reshape",
96
- "Div",
97
- "Slice",
98
- "Add",
99
  "Sub",
100
- "MatMul",
101
- "Constant",
102
  "Mul",
103
- "Range",
104
- "Less",
105
- "Gather",
106
- "Expand",
107
  "Sigmoid",
108
- "ConstantOfShape",
 
109
  "Equal",
110
- "Sqrt",
111
- "Cast",
112
  "Where",
113
- "If",
114
- "Unsqueeze",
115
- "Softmax",
116
- "Pow",
117
  "Transpose",
118
- "Concat",
119
- "Shape",
120
- "Squeeze"
 
 
 
 
 
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
 
2
  "per_channel": true,
3
  "reduce_range": true,
4
  "per_model_config": {
5
+ "decoder_model": {
6
  "op_types": [
7
+ "Sqrt",
8
+ "Shape",
 
 
9
  "Sub",
 
 
10
  "Mul",
11
+ "Less",
12
+ "Constant",
13
+ "Div",
14
  "Sigmoid",
15
+ "ReduceMean",
16
+ "Gather",
17
  "Equal",
18
+ "Concat",
19
+ "Slice",
20
+ "Range",
21
  "Where",
 
 
 
22
  "Transpose",
23
+ "Squeeze",
24
+ "Softmax",
25
+ "Expand",
26
+ "ConstantOfShape",
27
+ "Add",
28
+ "Reshape",
29
+ "MatMul",
30
+ "Unsqueeze",
31
+ "Cast",
32
+ "Pow"
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "encoder_model": {
37
  "op_types": [
38
+ "Sqrt",
39
+ "Shape",
 
40
  "Sub",
 
 
 
41
  "Mul",
42
+ "Div",
43
+ "Constant",
 
44
  "Sigmoid",
45
+ "ReduceMean",
46
+ "Gather",
47
  "Equal",
48
+ "Concat",
49
+ "Range",
50
  "Where",
 
 
 
51
  "Transpose",
52
+ "Softmax",
53
+ "Expand",
54
+ "ConstantOfShape",
55
+ "Add",
56
+ "Reshape",
57
+ "MatMul",
58
+ "Unsqueeze",
59
+ "Cast",
60
+ "Pow"
61
  ],
62
  "weight_type": "QInt8"
63
  },
64
+ "decoder_model_merged": {
65
  "op_types": [
66
+ "Sqrt",
67
+ "Shape",
 
 
 
68
  "Sub",
 
 
69
  "Mul",
 
70
  "Less",
71
+ "Constant",
72
+ "Div",
73
  "Sigmoid",
74
+ "If",
75
+ "ReduceMean",
76
+ "Gather",
77
  "Equal",
78
+ "Concat",
79
+ "Slice",
80
+ "Range",
81
  "Where",
 
 
 
82
  "Transpose",
83
+ "Squeeze",
84
+ "Softmax",
85
+ "Expand",
86
+ "ConstantOfShape",
87
+ "Add",
88
+ "Reshape",
89
+ "MatMul",
90
+ "Unsqueeze",
91
+ "Cast",
92
+ "Pow"
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
+ "decoder_with_past_model": {
97
  "op_types": [
98
+ "Sqrt",
99
+ "Shape",
 
 
 
100
  "Sub",
 
 
101
  "Mul",
102
+ "Div",
103
+ "Constant",
 
 
104
  "Sigmoid",
105
+ "ReduceMean",
106
+ "Gather",
107
  "Equal",
108
+ "Concat",
109
+ "Range",
110
  "Where",
 
 
 
 
111
  "Transpose",
112
+ "Softmax",
113
+ "Expand",
114
+ "ConstantOfShape",
115
+ "Add",
116
+ "Reshape",
117
+ "MatMul",
118
+ "Unsqueeze",
119
+ "Cast",
120
+ "Pow"
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff