Xenova HF staff commited on
Commit
d4ed587
1 Parent(s): 2a78872

Upload folder using huggingface_hub

Browse files
config.json ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "MIT/ast-finetuned-speech-commands-v2",
3
+ "architectures": [
4
+ "ASTForAudioClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "frequency_stride": 10,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.0,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "backward",
13
+ "1": "follow",
14
+ "2": "five",
15
+ "3": "bed",
16
+ "4": "zero",
17
+ "5": "on",
18
+ "6": "learn",
19
+ "7": "two",
20
+ "8": "house",
21
+ "9": "tree",
22
+ "10": "dog",
23
+ "11": "stop",
24
+ "12": "seven",
25
+ "13": "eight",
26
+ "14": "down",
27
+ "15": "six",
28
+ "16": "forward",
29
+ "17": "cat",
30
+ "18": "right",
31
+ "19": "visual",
32
+ "20": "four",
33
+ "21": "wow",
34
+ "22": "no",
35
+ "23": "nine",
36
+ "24": "off",
37
+ "25": "three",
38
+ "26": "left",
39
+ "27": "marvin",
40
+ "28": "yes",
41
+ "29": "up",
42
+ "30": "sheila",
43
+ "31": "happy",
44
+ "32": "bird",
45
+ "33": "go",
46
+ "34": "one"
47
+ },
48
+ "initializer_range": 0.02,
49
+ "intermediate_size": 3072,
50
+ "label2id": {
51
+ "backward": 0,
52
+ "bed": 3,
53
+ "bird": 32,
54
+ "cat": 17,
55
+ "dog": 10,
56
+ "down": 14,
57
+ "eight": 13,
58
+ "five": 2,
59
+ "follow": 1,
60
+ "forward": 16,
61
+ "four": 20,
62
+ "go": 33,
63
+ "happy": 31,
64
+ "house": 8,
65
+ "learn": 6,
66
+ "left": 26,
67
+ "marvin": 27,
68
+ "nine": 23,
69
+ "no": 22,
70
+ "off": 24,
71
+ "on": 5,
72
+ "one": 34,
73
+ "right": 18,
74
+ "seven": 12,
75
+ "sheila": 30,
76
+ "six": 15,
77
+ "stop": 11,
78
+ "three": 25,
79
+ "tree": 9,
80
+ "two": 7,
81
+ "up": 29,
82
+ "visual": 19,
83
+ "wow": 21,
84
+ "yes": 28,
85
+ "zero": 4
86
+ },
87
+ "layer_norm_eps": 1e-12,
88
+ "max_length": 128,
89
+ "model_type": "audio-spectrogram-transformer",
90
+ "num_attention_heads": 12,
91
+ "num_hidden_layers": 12,
92
+ "num_mel_bins": 128,
93
+ "patch_size": 16,
94
+ "qkv_bias": true,
95
+ "time_stride": 10,
96
+ "transformers_version": "4.30.2"
97
+ }
onnx/model.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3285c284ea4352c7289085c819e7b809e3eaca4da374f52ce136f9db3aadefa2
3
+ size 341960003
onnx/model_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81ec9acdcf903a2fc3ac1e389c7e7a040864a990f0c5fc8fd215ac68648b9cd1
3
+ size 87144196
preprocessor_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "ASTFeatureExtractor",
4
+ "feature_size": 1,
5
+ "max_length": 128,
6
+ "mean": -6.845978,
7
+ "num_mel_bins": 128,
8
+ "padding_side": "right",
9
+ "padding_value": 0.0,
10
+ "return_attention_mask": false,
11
+ "sampling_rate": 16000,
12
+ "std": 5.5654526
13
+ }
quant_config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": true,
3
+ "reduce_range": true,
4
+ "per_model_config": {
5
+ "model": {
6
+ "op_types": [
7
+ "Unsqueeze",
8
+ "Equal",
9
+ "Pow",
10
+ "Conv",
11
+ "Sub",
12
+ "Constant",
13
+ "Transpose",
14
+ "Shape",
15
+ "Slice",
16
+ "Where",
17
+ "Sqrt",
18
+ "Expand",
19
+ "Div",
20
+ "Gather",
21
+ "Mul",
22
+ "Softmax",
23
+ "MatMul",
24
+ "Concat",
25
+ "Add",
26
+ "Reshape",
27
+ "Gemm",
28
+ "ConstantOfShape",
29
+ "ReduceMean",
30
+ "Erf"
31
+ ],
32
+ "weight_type": "QUInt8"
33
+ }
34
+ }
35
+ }