Hugging Face's logo

Xenova
/
whisper-medium

Automatic Speech Recognition
Transformers.js
ONNX
whisper
Model card Files
xet
Community
1
Xenova HF Staff commited on
Commit
33bf94e
·
1 Parent(s): 60636d5

Upload folder using huggingface_hub

Browse files
Files changed (11) hide show
  1. config.json +2 -1
  2. generation_config.json +27 -2
  3. onnx/decoder_model.onnx +2 -2
  4. onnx/decoder_model_merged.onnx +2 -2
  5. onnx/decoder_model_merged_quantized.onnx +2 -2
  6. onnx/decoder_model_quantized.onnx +2 -2
  7. onnx/decoder_with_past_model.onnx +2 -2
  8. onnx/decoder_with_past_model_quantized.onnx +2 -2
  9. onnx/encoder_model.onnx +2 -2
  10. onnx/encoder_model_quantized.onnx +2 -2
  11. quantize_config.json +115 -0
config.json CHANGED
@@ -50,6 +50,7 @@
50
  "max_length": 448,
51
  "max_source_positions": 1500,
52
  "max_target_positions": 448,
 
53
  "model_type": "whisper",
54
  "num_hidden_layers": 24,
55
  "num_mel_bins": 80,
@@ -145,7 +146,7 @@
145
  50361,
146
  50362
147
  ],
148
- "transformers_version": "4.30.2",
149
  "use_cache": true,
150
  "use_weighted_layer_sum": false,
151
  "vocab_size": 51865
 
50
  "max_length": 448,
51
  "max_source_positions": 1500,
52
  "max_target_positions": 448,
53
+ "median_filter_width": 7,
54
  "model_type": "whisper",
55
  "num_hidden_layers": 24,
56
  "num_mel_bins": 80,
 
146
  50361,
147
  50362
148
  ],
149
+ "transformers_version": "4.33.0.dev0",
150
  "use_cache": true,
151
  "use_weighted_layer_sum": false,
152
  "vocab_size": 51865
generation_config.json CHANGED
@@ -1,5 +1,30 @@
1
  {
2
- "alignment_heads": [[13, 15], [15, 4], [15, 15], [16, 1], [20, 0], [23, 4]],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  "begin_suppress_tokens": [
4
  220,
5
  50257
@@ -217,5 +242,5 @@
217
  "transcribe": 50359,
218
  "translate": 50358
219
  },
220
- "transformers_version": "4.30.2"
221
  }
 
1
  {
2
+ "alignment_heads": [
3
+ [
4
+ 13,
5
+ 15
6
+ ],
7
+ [
8
+ 15,
9
+ 4
10
+ ],
11
+ [
12
+ 15,
13
+ 15
14
+ ],
15
+ [
16
+ 16,
17
+ 1
18
+ ],
19
+ [
20
+ 20,
21
+ 0
22
+ ],
23
+ [
24
+ 23,
25
+ 4
26
+ ]
27
+ ],
28
  "begin_suppress_tokens": [
29
  220,
30
  50257
 
242
  "transcribe": 50359,
243
  "translate": 50358
244
  },
245
+ "transformers_version": "4.33.0.dev0"
246
  }
onnx/decoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0a648963b5554dd3b637976d78be6befa6093f59dbfcb6374c0363acdb953c1a
3
- size 2040015092
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00ea982dc84ad2d2904b46749eab6f27aa56287d523ae4d5dbcd0f2fd5ae902b
3
+ size 1827707405
onnx/decoder_model_merged.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6f82a91e07ab23f138384c1d6aba1f20b3437ccb6beb3eaa21b4ef74cdc292ed
3
- size 1926573
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3b860eefa7321ea078f00a5d4b821f36dd38362ba3641a579ddfde3df37b45e
3
+ size 1828728265
onnx/decoder_model_merged_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0bdb36b4facd9a224dde4533588f876ab0fb3599001bde4869a39e014594ce13
3
- size 515407456
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2cdd6d06ebdf9d993d21117bfeeb7e9b399521b7766d3df77c54a85d6dcf3c08
3
+ size 462661606
onnx/decoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bf74fbc6d746531b96673d35e8bd455bb049a4f570847b925e9685fe426bc5a0
3
- size 514019282
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8aa8c0ac25486ecb9dc83cf49b08d6abd22d4ef6960f3d21c00a3ab6f1a8a439
3
+ size 461079957
onnx/decoder_with_past_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:885893286374cb2cf662a74c0912ff6c47f0caf80964153555a0398cbeca3b86
3
- size 1838441751
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6bd9e6475391368696a8bf1c8ba10d3a0f8c07ea7e3b3545ed7ccedb28526cd4
3
+ size 1626146694
onnx/decoder_with_past_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e81dd3275a51a8fd1a41247b7cfb4bba5df7d10b3814640c7920579bcca19ba6
3
- size 463340703
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d652797726d1662d3fe72daa93fccc814e42e68e119d827795ce712f17a705a8
3
+ size 410418679
onnx/encoder_model.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:70292a0b342decb984585a1dcd93ad204dc9640370af48a2f829220483b2835c
3
- size 1229258863
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2089510b7427f03f3126e35a007b9eaf64bd7497d37d13e763264e9178f4a267
3
+ size 1229312445
onnx/encoder_model_quantized.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5d01068c82a2d3e963ffe9047b5cb802539a4f26980434c00f844d431a794ff1
3
- size 313397176
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d6b4a00e441271646327f8a71b6e1bd1a305013cd914b51ddd76919c59ee3af
3
+ size 313468028
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
+ "Expand",
8
+ "Concat",
9
+ "Div",
10
+ "Cast",
11
+ "Slice",
12
+ "Mul",
13
+ "Where",
14
+ "Transpose",
15
+ "Erf",
16
+ "Reshape",
17
+ "Equal",
18
+ "ConstantOfShape",
19
+ "Unsqueeze",
20
+ "ReduceMean",
21
+ "Softmax",
22
+ "Squeeze",
23
+ "Less",
24
+ "Range",
25
+ "MatMul",
26
+ "Pow",
27
+ "Sqrt",
28
+ "Sub",
29
+ "Add",
30
+ "Constant",
31
+ "Gather",
32
+ "Shape"
33
+ ],
34
+ "weight_type": "QInt8"
35
+ },
36
+ "decoder_model_merged": {
37
+ "op_types": [
38
+ "Expand",
39
+ "Concat",
40
+ "Div",
41
+ "Cast",
42
+ "Slice",
43
+ "Mul",
44
+ "Where",
45
+ "Transpose",
46
+ "Erf",
47
+ "Reshape",
48
+ "Equal",
49
+ "ConstantOfShape",
50
+ "Unsqueeze",
51
+ "ReduceMean",
52
+ "Softmax",
53
+ "Squeeze",
54
+ "Less",
55
+ "Range",
56
+ "MatMul",
57
+ "Pow",
58
+ "If",
59
+ "Sqrt",
60
+ "Sub",
61
+ "Add",
62
+ "Constant",
63
+ "Gather",
64
+ "Shape"
65
+ ],
66
+ "weight_type": "QInt8"
67
+ },
68
+ "decoder_with_past_model": {
69
+ "op_types": [
70
+ "ReduceMean",
71
+ "Div",
72
+ "Concat",
73
+ "Sqrt",
74
+ "Sub",
75
+ "Softmax",
76
+ "Add",
77
+ "Transpose",
78
+ "Constant",
79
+ "Gather",
80
+ "Shape",
81
+ "Slice",
82
+ "Erf",
83
+ "Reshape",
84
+ "Mul",
85
+ "MatMul",
86
+ "Unsqueeze",
87
+ "Pow"
88
+ ],
89
+ "weight_type": "QInt8"
90
+ },
91
+ "encoder_model": {
92
+ "op_types": [
93
+ "ReduceMean",
94
+ "Div",
95
+ "Sqrt",
96
+ "Concat",
97
+ "Sub",
98
+ "Softmax",
99
+ "Add",
100
+ "Transpose",
101
+ "Constant",
102
+ "Gather",
103
+ "Shape",
104
+ "Erf",
105
+ "Reshape",
106
+ "Mul",
107
+ "Conv",
108
+ "MatMul",
109
+ "Unsqueeze",
110
+ "Pow"
111
+ ],
112
+ "weight_type": "QUInt8"
113
+ }
114
+ }
115
+ }