Xenova HF staff commited on
Commit
7f9967e
1 Parent(s): 1011e60

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -56,7 +56,7 @@
56
  "scale_embedding": true,
57
  "share_encoder_decoder_embeddings": true,
58
  "static_position_embeddings": true,
59
- "transformers_version": "4.32.0.dev0",
60
  "use_cache": true,
61
  "vocab_size": 65001
62
  }
 
56
  "scale_embedding": true,
57
  "share_encoder_decoder_embeddings": true,
58
  "static_position_embeddings": true,
59
+ "transformers_version": "4.34.0.dev0",
60
  "use_cache": true,
61
  "vocab_size": 65001
62
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  65000
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 4,
14
  "pad_token_id": 65000,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  65000
 
11
  "max_length": 512,
12
  "num_beams": 4,
13
  "pad_token_id": 65000,
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:ebd30a9a72623a01c4dae38421de4690f15935d9baf64ab14679b96bd04331ad
3
- size 368725660
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f88068f2e88da6f1e603265e0fdecd4498e7bfbed0c33c5e22ed2f126553cfd
3
+ size 235603669
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:550fb4152aae3cab263e73700536e5af3a17e6deb665d8a0eeee0440504c6140
3
- size 368961142
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:543a46ff7d2e45897770f24df97d21a8b0a50cf312f5ee3754df587c1883e957
3
+ size 235839236
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6bb8b744f6cda1057e326fdae4566d5a8d6fb2bc2e0d2ab935a106eea2edadca
3
- size 93819733
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58ee5ddd6e22d1693d8722b90c1486afb93700a4376cf28fd74175052ad530ae
3
+ size 60212804
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a4c423621605ae0e6587d8c0324df4e3f8a2c9497a129360bbfef23f2e1fe8ca
3
- size 93448377
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b45074cc819f1d901e40510d7bc0781d99327d16c7eca04518c4ee0c3d6ba10
3
+ size 59842102
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:791ed805f79bec0f0ab50ba6619f47a9c6b4b2e4ee44ef3151cca4e212bf5659
3
- size 356075151
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba0c5f52a41fe46c3267a587f0102e0ad94260d3f692dfd921d8b94568479af4
3
+ size 222953160
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f9bc8a01dfd322a842e7c1ca0bf0f682aa4a16c7dfc75ce9169909081da6ab85
3
- size 90176913
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:163a130a2cc61d1326030abe338af30f3c9b864a2c5405de11648e7ccb1c1dac
3
+ size 56570638
quantize_config.json CHANGED
@@ -4,120 +4,120 @@
4
  "per_model_config": {
5
  "encoder_model": {
6
  "op_types": [
 
 
 
 
 
 
7
  "Reshape",
8
- "Transpose",
9
- "Constant",
10
  "Sub",
11
- "MatMul",
12
- "ReduceMean",
13
  "Pow",
 
 
 
14
  "Gather",
15
- "Equal",
16
- "Unsqueeze",
17
  "Sigmoid",
18
- "Cast",
19
- "Where",
20
- "Expand",
21
- "Concat",
22
- "Div",
23
- "Softmax",
24
- "Mul",
25
- "Sqrt",
26
  "Shape",
27
- "ConstantOfShape",
28
  "Add",
29
- "Range"
 
 
 
 
 
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
  "decoder_with_past_model": {
34
  "op_types": [
 
 
 
 
 
 
35
  "Reshape",
36
- "Transpose",
37
- "Constant",
38
  "Sub",
39
- "MatMul",
40
- "ReduceMean",
41
  "Pow",
 
 
 
42
  "Gather",
43
- "Equal",
44
- "Unsqueeze",
45
  "Sigmoid",
46
- "Where",
47
- "Cast",
48
- "Expand",
49
- "Concat",
50
- "Div",
51
- "Softmax",
52
- "Mul",
53
- "Sqrt",
54
  "Shape",
55
- "ConstantOfShape",
56
  "Add",
57
- "Range"
 
 
 
 
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
 
 
 
 
 
 
 
63
  "Reshape",
64
- "Transpose",
65
- "Constant",
66
  "Sub",
67
- "MatMul",
68
- "ReduceMean",
69
  "Pow",
70
- "Gather",
71
- "Equal",
72
- "Unsqueeze",
73
- "Sigmoid",
74
- "Cast",
75
- "Where",
76
- "Expand",
77
  "Concat",
78
- "Div",
79
- "Softmax",
80
- "Mul",
81
- "Less",
82
  "Slice",
83
- "Sqrt",
84
  "Shape",
85
- "ConstantOfShape",
86
  "Add",
87
- "Range",
88
- "Squeeze"
 
 
 
 
 
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
 
 
 
 
 
 
 
94
  "Reshape",
95
- "Transpose",
96
- "Constant",
97
  "Sub",
98
- "MatMul",
99
- "ReduceMean",
100
  "Pow",
101
- "Gather",
102
- "Equal",
103
- "Unsqueeze",
104
- "Sigmoid",
105
- "Cast",
106
- "Where",
107
- "Expand",
108
  "Concat",
109
- "Div",
110
- "Softmax",
111
- "Mul",
112
- "Less",
113
  "Slice",
114
- "Sqrt",
115
  "Shape",
116
- "If",
117
- "ConstantOfShape",
118
  "Add",
119
- "Range",
120
- "Squeeze"
 
 
 
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
 
4
  "per_model_config": {
5
  "encoder_model": {
6
  "op_types": [
7
+ "Equal",
8
+ "Range",
9
+ "Div",
10
+ "ReduceMean",
11
+ "Sqrt",
12
+ "MatMul",
13
  "Reshape",
14
+ "Unsqueeze",
 
15
  "Sub",
 
 
16
  "Pow",
17
+ "Transpose",
18
+ "ConstantOfShape",
19
+ "Concat",
20
  "Gather",
 
 
21
  "Sigmoid",
 
 
 
 
 
 
 
 
22
  "Shape",
 
23
  "Add",
24
+ "Expand",
25
+ "Mul",
26
+ "Softmax",
27
+ "Where",
28
+ "Constant",
29
+ "Cast"
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
  "decoder_with_past_model": {
34
  "op_types": [
35
+ "Equal",
36
+ "Range",
37
+ "Div",
38
+ "ReduceMean",
39
+ "Sqrt",
40
+ "MatMul",
41
  "Reshape",
42
+ "Unsqueeze",
 
43
  "Sub",
 
 
44
  "Pow",
45
+ "Transpose",
46
+ "ConstantOfShape",
47
+ "Concat",
48
  "Gather",
 
 
49
  "Sigmoid",
 
 
 
 
 
 
 
 
50
  "Shape",
51
+ "Expand",
52
  "Add",
53
+ "Mul",
54
+ "Softmax",
55
+ "Where",
56
+ "Constant",
57
+ "Cast"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
+ "decoder_model_merged": {
62
  "op_types": [
63
+ "Equal",
64
+ "Range",
65
+ "Less",
66
+ "Div",
67
+ "ReduceMean",
68
+ "Sqrt",
69
+ "MatMul",
70
  "Reshape",
71
+ "Unsqueeze",
 
72
  "Sub",
 
 
73
  "Pow",
74
+ "Squeeze",
75
+ "Transpose",
76
+ "ConstantOfShape",
 
 
 
 
77
  "Concat",
78
+ "Gather",
 
 
 
79
  "Slice",
80
+ "Sigmoid",
81
  "Shape",
 
82
  "Add",
83
+ "Expand",
84
+ "Mul",
85
+ "Softmax",
86
+ "If",
87
+ "Constant",
88
+ "Where",
89
+ "Cast"
90
  ],
91
  "weight_type": "QInt8"
92
  },
93
+ "decoder_model": {
94
  "op_types": [
95
+ "Equal",
96
+ "Range",
97
+ "Less",
98
+ "Div",
99
+ "ReduceMean",
100
+ "Sqrt",
101
+ "MatMul",
102
  "Reshape",
103
+ "Unsqueeze",
 
104
  "Sub",
 
 
105
  "Pow",
106
+ "Squeeze",
107
+ "Transpose",
108
+ "ConstantOfShape",
 
 
 
 
109
  "Concat",
110
+ "Gather",
 
 
 
111
  "Slice",
112
+ "Sigmoid",
113
  "Shape",
 
 
114
  "Add",
115
+ "Expand",
116
+ "Mul",
117
+ "Softmax",
118
+ "Where",
119
+ "Constant",
120
+ "Cast"
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff