Xenova HF staff commited on
Commit
361b5c9
1 Parent(s): e80686a

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -139,8 +139,7 @@
139
  50360,
140
  50361
141
  ],
142
- "torch_dtype": "float32",
143
- "transformers_version": "4.29.2",
144
  "use_cache": true,
145
  "use_weighted_layer_sum": false,
146
  "vocab_size": 51864
 
139
  50360,
140
  50361
141
  ],
142
+ "transformers_version": "4.30.2",
 
143
  "use_cache": true,
144
  "use_weighted_layer_sum": false,
145
  "vocab_size": 51864
generation_config.json CHANGED
@@ -109,5 +109,5 @@
109
  50360,
110
  50361
111
  ],
112
- "transformers_version": "4.29.2"
113
  }
 
109
  50360,
110
  50361
111
  ],
112
+ "transformers_version": "4.30.2"
113
  }
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2746e6eee2fc3dacc4ee3aa7024782755da6b3b4b351aad9ed573d14682e503a
3
  size 774588347
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:821a0fae52d92575d988421df91cc2792827b58f42d6274324fa184a592dc6a2
3
  size 774588347
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:38ddd300ced7653406425d69833e2d7901f36edd4767d2663896b026d467e90d
3
- size 197289993
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba8567520a70349d064d7ffec5f440e78a149bf5968bfedd94ad78b6287e8798
3
+ size 196431022
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5d44e700708c59e0498b8b986a817f69ce311e25a9cd01df4f9b5ebf4f118a10
3
- size 196596602
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:053b11d598b22f489e0075714ca9191f4016c356544131b908f2eb87648d72c2
3
+ size 195737631
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2a2b697c212ac0e55fce6fcbfd0053ac1f2253c2394d881479529dec251d0767
3
- size 182183920
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5b511413b04bb36148c2b45799223c14d5075bd6b1f16d16145d14fb6adc57f
3
+ size 181417229
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f25862e6297bed816b7ec75f51b0c1f937e312e1032362b579f4cd6e1c8b4395
3
- size 92704740
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc7c82e4908c040e156b4063f6cc319b5f31d68ff5812a7e4ac05659cb88ac5c
3
+ size 92289703
quant_config.json ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": false,
3
+ "reduce_range": false,
4
+ "per_model_config": {
5
+ "encoder_model": {
6
+ "op_types": [
7
+ "Add",
8
+ "Gather",
9
+ "Mul",
10
+ "Pow",
11
+ "Sub",
12
+ "Conv",
13
+ "MatMul",
14
+ "Erf",
15
+ "Transpose",
16
+ "Sqrt",
17
+ "Constant",
18
+ "Shape",
19
+ "Softmax",
20
+ "Unsqueeze",
21
+ "Div",
22
+ "Concat",
23
+ "Reshape",
24
+ "ReduceMean"
25
+ ],
26
+ "weight_type": "QUInt8"
27
+ },
28
+ "decoder_model": {
29
+ "op_types": [
30
+ "MatMul",
31
+ "Squeeze",
32
+ "Shape",
33
+ "Div",
34
+ "Reshape",
35
+ "ReduceMean",
36
+ "Gather",
37
+ "Cast",
38
+ "Mul",
39
+ "Expand",
40
+ "Pow",
41
+ "Erf",
42
+ "Slice",
43
+ "Transpose",
44
+ "Constant",
45
+ "Softmax",
46
+ "Unsqueeze",
47
+ "Concat",
48
+ "ConstantOfShape",
49
+ "Equal",
50
+ "Where",
51
+ "Range",
52
+ "Sub",
53
+ "Less",
54
+ "Add",
55
+ "Sqrt"
56
+ ],
57
+ "weight_type": "QInt8"
58
+ },
59
+ "decoder_model_merged": {
60
+ "op_types": [
61
+ "If"
62
+ ],
63
+ "weight_type": "QInt8"
64
+ },
65
+ "decoder_with_past_model": {
66
+ "op_types": [
67
+ "Gather",
68
+ "Add",
69
+ "Mul",
70
+ "Pow",
71
+ "Sub",
72
+ "MatMul",
73
+ "Erf",
74
+ "Slice",
75
+ "Transpose",
76
+ "Sqrt",
77
+ "Constant",
78
+ "Shape",
79
+ "Softmax",
80
+ "Unsqueeze",
81
+ "Concat",
82
+ "Div",
83
+ "Reshape",
84
+ "ReduceMean"
85
+ ],
86
+ "weight_type": "QInt8"
87
+ }
88
+ }
89
+ }