Hugging Face's logo

mesolitica
/
whisper-base-isclean

Audio Classification
Transformers
Safetensors
whisper
Inference Endpoints
Model card Files Community
aisyahhrazak commited on
Commit
8d55854
•
1 Parent(s): f7921ff

Upload WhisperForAudioClassification

Browse files
Files changed (2) hide show
  1. config.json +162 -0
  2. model.safetensors +3 -0
config.json ADDED
@@ -0,0 +1,162 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "whisper-base-is-clean/checkpoint-22000",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": false,
6
+ "architectures": [
7
+ "WhisperForAudioClassification"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": [
11
+ 220,
12
+ 50257
13
+ ],
14
+ "bos_token_id": 50257,
15
+ "classifier_proj_size": 256,
16
+ "d_model": 512,
17
+ "decoder_attention_heads": 8,
18
+ "decoder_ffn_dim": 2048,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 6,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 8,
24
+ "encoder_ffn_dim": 2048,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 6,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": [
29
+ [
30
+ 1,
31
+ 50259
32
+ ],
33
+ [
34
+ 2,
35
+ 50359
36
+ ],
37
+ [
38
+ 3,
39
+ 50363
40
+ ]
41
+ ],
42
+ "id2label": {
43
+ "0": "noise",
44
+ "1": "clean"
45
+ },
46
+ "init_std": 0.02,
47
+ "is_encoder_decoder": true,
48
+ "label2id": {
49
+ "clean": 1,
50
+ "noise": 0
51
+ },
52
+ "mask_feature_length": 10,
53
+ "mask_feature_min_masks": 0,
54
+ "mask_feature_prob": 0.0,
55
+ "mask_time_length": 10,
56
+ "mask_time_min_masks": 2,
57
+ "mask_time_prob": 0.05,
58
+ "max_length": 448,
59
+ "max_source_positions": 1500,
60
+ "max_target_positions": 448,
61
+ "median_filter_width": 7,
62
+ "model_type": "whisper",
63
+ "num_hidden_layers": 6,
64
+ "num_mel_bins": 80,
65
+ "pad_token_id": 50257,
66
+ "scale_embedding": false,
67
+ "suppress_tokens": [
68
+ 1,
69
+ 2,
70
+ 7,
71
+ 8,
72
+ 9,
73
+ 10,
74
+ 14,
75
+ 25,
76
+ 26,
77
+ 27,
78
+ 28,
79
+ 29,
80
+ 31,
81
+ 58,
82
+ 59,
83
+ 60,
84
+ 61,
85
+ 62,
86
+ 63,
87
+ 90,
88
+ 91,
89
+ 92,
90
+ 93,
91
+ 359,
92
+ 503,
93
+ 522,
94
+ 542,
95
+ 873,
96
+ 893,
97
+ 902,
98
+ 918,
99
+ 922,
100
+ 931,
101
+ 1350,
102
+ 1853,
103
+ 1982,
104
+ 2460,
105
+ 2627,
106
+ 3246,
107
+ 3253,
108
+ 3268,
109
+ 3536,
110
+ 3846,
111
+ 3961,
112
+ 4183,
113
+ 4667,
114
+ 6585,
115
+ 6647,
116
+ 7273,
117
+ 9061,
118
+ 9383,
119
+ 10428,
120
+ 10929,
121
+ 11938,
122
+ 12033,
123
+ 12331,
124
+ 12562,
125
+ 13793,
126
+ 14157,
127
+ 14635,
128
+ 15265,
129
+ 15618,
130
+ 16553,
131
+ 16604,
132
+ 18362,
133
+ 18956,
134
+ 20075,
135
+ 21675,
136
+ 22520,
137
+ 26130,
138
+ 26161,
139
+ 26435,
140
+ 28279,
141
+ 29464,
142
+ 31650,
143
+ 32302,
144
+ 32470,
145
+ 36865,
146
+ 42863,
147
+ 47425,
148
+ 49870,
149
+ 50254,
150
+ 50258,
151
+ 50358,
152
+ 50359,
153
+ 50360,
154
+ 50361,
155
+ 50362
156
+ ],
157
+ "torch_dtype": "float32",
158
+ "transformers_version": "4.36.1",
159
+ "use_cache": true,
160
+ "use_weighted_layer_sum": false,
161
+ "vocab_size": 51865
162
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6ccab75bf5152ba9064956ba0e1eb74b53175c968e20527456ef9945a1a3eec
3
+ size 82900264