Xenova HF staff commited on
Commit
d88e1c8
1 Parent(s): 3ccf208

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": 62834
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": 62834
60
  }
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "bad_words_ids": [
4
  [
5
  62833
@@ -12,5 +11,6 @@
12
  "max_length": 512,
13
  "num_beams": 4,
14
  "pad_token_id": 62833,
15
- "transformers_version": "4.32.0.dev0"
 
16
  }
 
1
  {
 
2
  "bad_words_ids": [
3
  [
4
  62833
 
11
  "max_length": 512,
12
  "num_beams": 4,
13
  "pad_token_id": 62833,
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 sha256:125b42cc04865b0ff6e64b247a845e45ae1a4b0d45db72d14f1963fbe4508522
3
- size 359840960
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5623945b6332e8785e9a8a7e83298b8c00803e4b3285e267b15265b0ab71b0f
3
+ size 231156985
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d67cfd1baf59398a90885ae42e9a4f1cd7cf2c4e2c6a2eb82f2aaab694762b0a
3
- size 360076442
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0113ffe211c4d84afbb7f537900aa47b526dc276bfb57826458040515578f18
3
+ size 231392552
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e754c5f68327792694e2cf073e5de0b790de019de467017ec44cc31bff17d12d
3
- size 91581222
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cdb25ae736afe9d48035b0ea311ab4b25f315810fab7f3c6f6cb0d8d87af35fa
3
+ size 59094632
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6f15ee20988afb1e97c8557a91744d3c41c03acf3c91f1319d0a522e1def20f3
3
- size 91209866
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:66f2be0af1ad3d875315f7e8ad2de4f8248fba75dc7173b1a93bf48023041e80
3
+ size 58723930
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1c55d8d96d15be9a0870ddd7e46702a238fd3fd011c46c72ba2c37268b22d95
3
- size 347190451
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f789c42d29a7b90103b0f200b99a453b8444a34e83dc1082bc69dbb2c57c7d06
3
+ size 218506476
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a3c44860e389b88e49c401ec26b70667038d46f2f184efe49fcafc43a3753d46
3
- size 87938402
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b69d795a40b8f645a980b458cc554bf739ac82bfd05345970a3756f66015293b
3
+ size 55452466
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
- "Mul",
8
- "MatMul",
9
- "Cast",
10
  "Pow",
11
- "Sub",
12
- "Sigmoid",
13
- "ReduceMean",
14
- "Where",
15
- "Transpose",
16
- "Unsqueeze",
17
- "Expand",
18
  "Sqrt",
19
- "Add",
 
 
 
20
  "Constant",
 
 
 
 
 
21
  "Shape",
 
 
22
  "Concat",
 
 
23
  "Reshape",
24
- "Div",
25
- "Equal",
26
- "Range",
27
- "ConstantOfShape",
28
- "Softmax",
29
- "Gather"
30
  ],
31
  "weight_type": "QInt8"
32
  },
33
- "decoder_with_past_model": {
34
  "op_types": [
35
- "Mul",
36
- "MatMul",
37
- "Cast",
38
  "Pow",
39
- "Sub",
40
- "Sigmoid",
41
- "ReduceMean",
42
- "Where",
43
- "Transpose",
44
- "Unsqueeze",
45
- "Expand",
46
  "Sqrt",
47
- "Add",
 
 
 
48
  "Constant",
 
 
 
 
49
  "Shape",
 
 
50
  "Concat",
 
 
51
  "Reshape",
52
- "Div",
53
- "Equal",
54
- "Range",
55
- "ConstantOfShape",
56
- "Softmax",
57
- "Gather"
58
  ],
59
  "weight_type": "QInt8"
60
  },
61
- "decoder_model": {
62
  "op_types": [
63
- "Mul",
64
- "MatMul",
65
- "Cast",
66
  "Pow",
67
- "Sub",
68
- "Sigmoid",
69
- "ReduceMean",
70
- "Less",
71
- "Where",
72
- "Slice",
73
- "Transpose",
74
- "Unsqueeze",
75
- "Expand",
76
  "Sqrt",
77
- "Squeeze",
78
- "Add",
 
 
 
79
  "Constant",
 
 
 
 
 
80
  "Shape",
 
 
81
  "Concat",
 
 
82
  "Reshape",
83
- "Div",
84
- "Equal",
85
- "Range",
86
- "ConstantOfShape",
87
- "Softmax",
88
- "Gather"
89
  ],
90
  "weight_type": "QInt8"
91
  },
92
- "decoder_model_merged": {
93
  "op_types": [
94
- "Mul",
95
- "MatMul",
96
- "Cast",
97
  "Pow",
98
- "Sub",
99
- "Sigmoid",
100
- "ReduceMean",
101
- "Less",
102
- "Where",
103
- "Slice",
104
- "Transpose",
105
- "Unsqueeze",
106
- "Expand",
107
  "Sqrt",
108
- "Squeeze",
109
- "Add",
110
- "If",
 
111
  "Constant",
 
 
 
 
112
  "Shape",
 
 
113
  "Concat",
 
 
114
  "Reshape",
115
- "Div",
116
- "Equal",
117
- "Range",
118
- "ConstantOfShape",
119
- "Softmax",
120
- "Gather"
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
+ "ConstantOfShape",
 
 
8
  "Pow",
 
 
 
 
 
 
 
9
  "Sqrt",
10
+ "MatMul",
11
+ "Softmax",
12
+ "Range",
13
+ "Gather",
14
  "Constant",
15
+ "Where",
16
+ "Mul",
17
+ "Add",
18
+ "Div",
19
+ "Less",
20
  "Shape",
21
+ "ReduceMean",
22
+ "Expand",
23
  "Concat",
24
+ "Cast",
25
+ "Transpose",
26
  "Reshape",
27
+ "Unsqueeze",
28
+ "Squeeze",
29
+ "Sigmoid",
30
+ "Slice",
31
+ "Sub",
32
+ "Equal"
33
  ],
34
  "weight_type": "QInt8"
35
  },
36
+ "encoder_model": {
37
  "op_types": [
38
+ "ConstantOfShape",
 
 
39
  "Pow",
 
 
 
 
 
 
 
40
  "Sqrt",
41
+ "MatMul",
42
+ "Softmax",
43
+ "Range",
44
+ "Gather",
45
  "Constant",
46
+ "Where",
47
+ "Mul",
48
+ "Add",
49
+ "Div",
50
  "Shape",
51
+ "ReduceMean",
52
+ "Expand",
53
  "Concat",
54
+ "Cast",
55
+ "Transpose",
56
  "Reshape",
57
+ "Unsqueeze",
58
+ "Sigmoid",
59
+ "Sub",
60
+ "Equal"
 
 
61
  ],
62
  "weight_type": "QInt8"
63
  },
64
+ "decoder_model_merged": {
65
  "op_types": [
66
+ "ConstantOfShape",
 
 
67
  "Pow",
 
 
 
 
 
 
 
 
 
68
  "Sqrt",
69
+ "MatMul",
70
+ "Softmax",
71
+ "Range",
72
+ "If",
73
+ "Gather",
74
  "Constant",
75
+ "Where",
76
+ "Mul",
77
+ "Add",
78
+ "Div",
79
+ "Less",
80
  "Shape",
81
+ "ReduceMean",
82
+ "Expand",
83
  "Concat",
84
+ "Cast",
85
+ "Transpose",
86
  "Reshape",
87
+ "Unsqueeze",
88
+ "Squeeze",
89
+ "Sigmoid",
90
+ "Slice",
91
+ "Sub",
92
+ "Equal"
93
  ],
94
  "weight_type": "QInt8"
95
  },
96
+ "decoder_with_past_model": {
97
  "op_types": [
98
+ "ConstantOfShape",
 
 
99
  "Pow",
 
 
 
 
 
 
 
 
 
100
  "Sqrt",
101
+ "MatMul",
102
+ "Softmax",
103
+ "Range",
104
+ "Gather",
105
  "Constant",
106
+ "Where",
107
+ "Mul",
108
+ "Add",
109
+ "Div",
110
  "Shape",
111
+ "ReduceMean",
112
+ "Expand",
113
  "Concat",
114
+ "Cast",
115
+ "Transpose",
116
  "Reshape",
117
+ "Unsqueeze",
118
+ "Sigmoid",
119
+ "Sub",
120
+ "Equal"
 
 
121
  ],
122
  "weight_type": "QInt8"
123
  }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff