Xenova HF staff commited on
Commit
30ebb73
1 Parent(s): 9f92bc9

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -42,6 +42,7 @@
42
  "max_length": 448,
43
  "max_source_positions": 1500,
44
  "max_target_positions": 448,
 
45
  "model_type": "whisper",
46
  "num_hidden_layers": 4,
47
  "num_mel_bins": 80,
@@ -139,7 +140,7 @@
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
 
42
  "max_length": 448,
43
  "max_source_positions": 1500,
44
  "max_target_positions": 448,
45
+ "median_filter_width": 7,
46
  "model_type": "whisper",
47
  "num_hidden_layers": 4,
48
  "num_mel_bins": 80,
 
140
  50360,
141
  50361
142
  ],
143
+ "transformers_version": "4.33.0.dev0",
144
  "use_cache": true,
145
  "use_weighted_layer_sum": false,
146
  "vocab_size": 51864
generation_config.json CHANGED
@@ -1,5 +1,38 @@
1
  {
2
- "alignment_heads": [[1, 0], [2, 0], [2, 5], [3, 0], [3, 1], [3, 2], [3, 3], [3, 4]],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  "begin_suppress_tokens": [
4
  220,
5
  50256
@@ -111,5 +144,5 @@
111
  50360,
112
  50361
113
  ],
114
- "transformers_version": "4.30.2"
115
  }
 
1
  {
2
+ "alignment_heads": [
3
+ [
4
+ 1,
5
+ 0
6
+ ],
7
+ [
8
+ 2,
9
+ 0
10
+ ],
11
+ [
12
+ 2,
13
+ 5
14
+ ],
15
+ [
16
+ 3,
17
+ 0
18
+ ],
19
+ [
20
+ 3,
21
+ 1
22
+ ],
23
+ [
24
+ 3,
25
+ 2
26
+ ],
27
+ [
28
+ 3,
29
+ 3
30
+ ],
31
+ [
32
+ 3,
33
+ 4
34
+ ]
35
+ ],
36
  "begin_suppress_tokens": [
37
  220,
38
  50256
 
144
  50360,
145
  50361
146
  ],
147
+ "transformers_version": "4.33.0.dev0"
148
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f8bec5100bcac885428dc96253d46df583260bac7ec5c99a50897de8d6eb7f73
3
- size 198046458
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38f0b91131d1f1c8b2b5f1d1096f61e44ded18be49313c80abc5c11daf03bd28
3
+ size 118405120
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:441da7a85e2cac5508fad6795cf115333b8566cffc886326788e0a64366411a5
3
- size 198194466
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:950978b1dbcbf250335358c1236053ba19a7f7849b33dc777f4421b72b7626fa
3
+ size 118578063
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:29e4ab70caaedbfd18a4cc4fab3d0708b9eb7c888f54f31f11949a6a040fcf36
3
- size 50583569
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbb2e063b7fbc41d9803b9698f93ecb035c50cbb3fb87b56cb131e4a5eb99059
3
+ size 30727382
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:699a1478cd7b1126a8f345281c1fb29fa72a18b8f45e4b31c6499653e3dee1d2
3
- size 50347713
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62b1761a6c23961cdabbf7caa25495525d78528a64baacf7e4283c83cb2941f1
3
+ size 30459718
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dca3143327bca6dacbb59f4858e7a4901d0ac719cf468ce48a8e21494f8f1cca
3
- size 193292243
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52b796cccc711df396afa098ae7cb6271d7fad42e4e7188564448cfa99aed830
3
+ size 113653015
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1182513a4c3ccd16fbbabb2a760f86eac2e827d9f8ce4c78c33e3935af86526c
3
- size 49113801
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c7390ddd38c567a51a73f175c6655afcbcd5cf1d5dc6be5badc89cba3f1d59c
3
+ size 29228736
onnx/encoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:530ff754d57158293aea418249399a20af813593005195f84e16ab54150e5589
3
- size 32900723
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a048dcf0cde98db805f46be32b75d778cf824aad20b51a02e5b9cff457426238
3
+ size 32909539
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:66bb0bf174a3a3bcfdaccea161c811d0d90a25c927021ab9ffbf639b902225f1
3
- size 10113251
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cc3c6f8563d1b3fbd2c5af9f64c2bed8b020bc593c402d1ef53b9f08fbf1b90
3
+ size 10124913
quantize_config.json ADDED
@@ -0,0 +1,115 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": false,
3
+ "reduce_range": false,
4
+ "per_model_config": {
5
+ "decoder_model": {
6
+ "op_types": [
7
+ "ReduceMean",
8
+ "Concat",
9
+ "Sqrt",
10
+ "Less",
11
+ "Unsqueeze",
12
+ "Add",
13
+ "Cast",
14
+ "Div",
15
+ "Equal",
16
+ "Softmax",
17
+ "Pow",
18
+ "Squeeze",
19
+ "Where",
20
+ "Sub",
21
+ "ConstantOfShape",
22
+ "Erf",
23
+ "Transpose",
24
+ "Reshape",
25
+ "MatMul",
26
+ "Gather",
27
+ "Shape",
28
+ "Mul",
29
+ "Range",
30
+ "Constant",
31
+ "Slice",
32
+ "Expand"
33
+ ],
34
+ "weight_type": "QInt8"
35
+ },
36
+ "decoder_model_merged": {
37
+ "op_types": [
38
+ "ReduceMean",
39
+ "Concat",
40
+ "Sqrt",
41
+ "Less",
42
+ "If",
43
+ "Unsqueeze",
44
+ "Add",
45
+ "Cast",
46
+ "Div",
47
+ "Equal",
48
+ "Softmax",
49
+ "Pow",
50
+ "Squeeze",
51
+ "Where",
52
+ "Sub",
53
+ "ConstantOfShape",
54
+ "Erf",
55
+ "Transpose",
56
+ "Reshape",
57
+ "MatMul",
58
+ "Gather",
59
+ "Shape",
60
+ "Mul",
61
+ "Range",
62
+ "Constant",
63
+ "Slice",
64
+ "Expand"
65
+ ],
66
+ "weight_type": "QInt8"
67
+ },
68
+ "decoder_with_past_model": {
69
+ "op_types": [
70
+ "MatMul",
71
+ "Gather",
72
+ "ReduceMean",
73
+ "Sub",
74
+ "Div",
75
+ "Concat",
76
+ "Sqrt",
77
+ "Shape",
78
+ "Softmax",
79
+ "Pow",
80
+ "Erf",
81
+ "Transpose",
82
+ "Reshape",
83
+ "Mul",
84
+ "Constant",
85
+ "Unsqueeze",
86
+ "Add",
87
+ "Slice"
88
+ ],
89
+ "weight_type": "QInt8"
90
+ },
91
+ "encoder_model": {
92
+ "op_types": [
93
+ "MatMul",
94
+ "ReduceMean",
95
+ "Conv",
96
+ "Div",
97
+ "Sub",
98
+ "Sqrt",
99
+ "Gather",
100
+ "Shape",
101
+ "Concat",
102
+ "Softmax",
103
+ "Pow",
104
+ "Erf",
105
+ "Transpose",
106
+ "Reshape",
107
+ "Mul",
108
+ "Constant",
109
+ "Unsqueeze",
110
+ "Add"
111
+ ],
112
+ "weight_type": "QUInt8"
113
+ }
114
+ }
115
+ }