Hugging Face's logo

sharkeyboi
/
ast-finetuned-audioset-10-10-0.4593-finetuned-frog_dataset

Audio Classification
Transformers
Safetensors
audio-spectrogram-transformer
Inference Endpoints
Model card Files Community
sharkeyboi commited on
Commit
d58660b
•
1 Parent(s): 3301fee

Training in progress, epoch 1

Browse files
Files changed (4) hide show
  1. config.json +201 -135
  2. model.safetensors +2 -2
  3. preprocessor_config.json +13 -13
  4. training_args.bin +1 -1
config.json CHANGED
@@ -1,135 +1,201 @@
1
- {
2
- "_name_or_path": "MIT/ast-finetuned-audioset-10-10-0.4593",
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": "LABEL_0",
13
- "1": "LABEL_1",
14
- "2": "LABEL_2",
15
- "3": "LABEL_3",
16
- "4": "LABEL_4",
17
- "5": "LABEL_5",
18
- "6": "LABEL_6",
19
- "7": "LABEL_7",
20
- "8": "LABEL_8",
21
- "9": "LABEL_9",
22
- "10": "LABEL_10",
23
- "11": "LABEL_11",
24
- "12": "LABEL_12",
25
- "13": "LABEL_13",
26
- "14": "LABEL_14",
27
- "15": "LABEL_15",
28
- "16": "LABEL_16",
29
- "17": "LABEL_17",
30
- "18": "LABEL_18",
31
- "19": "LABEL_19",
32
- "20": "LABEL_20",
33
- "21": "LABEL_21",
34
- "22": "LABEL_22",
35
- "23": "LABEL_23",
36
- "24": "LABEL_24",
37
- "25": "LABEL_25",
38
- "26": "LABEL_26",
39
- "27": "LABEL_27",
40
- "28": "LABEL_28",
41
- "29": "LABEL_29",
42
- "30": "LABEL_30",
43
- "31": "LABEL_31",
44
- "32": "LABEL_32",
45
- "33": "LABEL_33",
46
- "34": "LABEL_34",
47
- "35": "LABEL_35",
48
- "36": "LABEL_36",
49
- "37": "LABEL_37",
50
- "38": "LABEL_38",
51
- "39": "LABEL_39",
52
- "40": "LABEL_40",
53
- "41": "LABEL_41",
54
- "42": "LABEL_42",
55
- "43": "LABEL_43",
56
- "44": "LABEL_44",
57
- "45": "LABEL_45",
58
- "46": "LABEL_46",
59
- "47": "LABEL_47",
60
- "48": "LABEL_48",
61
- "49": "LABEL_49",
62
- "50": "LABEL_50",
63
- "51": "LABEL_51",
64
- "52": "LABEL_52"
65
- },
66
- "initializer_range": 0.02,
67
- "intermediate_size": 3072,
68
- "label2id": {
69
- "LABEL_0": 0,
70
- "LABEL_1": 1,
71
- "LABEL_10": 10,
72
- "LABEL_11": 11,
73
- "LABEL_12": 12,
74
- "LABEL_13": 13,
75
- "LABEL_14": 14,
76
- "LABEL_15": 15,
77
- "LABEL_16": 16,
78
- "LABEL_17": 17,
79
- "LABEL_18": 18,
80
- "LABEL_19": 19,
81
- "LABEL_2": 2,
82
- "LABEL_20": 20,
83
- "LABEL_21": 21,
84
- "LABEL_22": 22,
85
- "LABEL_23": 23,
86
- "LABEL_24": 24,
87
- "LABEL_25": 25,
88
- "LABEL_26": 26,
89
- "LABEL_27": 27,
90
- "LABEL_28": 28,
91
- "LABEL_29": 29,
92
- "LABEL_3": 3,
93
- "LABEL_30": 30,
94
- "LABEL_31": 31,
95
- "LABEL_32": 32,
96
- "LABEL_33": 33,
97
- "LABEL_34": 34,
98
- "LABEL_35": 35,
99
- "LABEL_36": 36,
100
- "LABEL_37": 37,
101
- "LABEL_38": 38,
102
- "LABEL_39": 39,
103
- "LABEL_4": 4,
104
- "LABEL_40": 40,
105
- "LABEL_41": 41,
106
- "LABEL_42": 42,
107
- "LABEL_43": 43,
108
- "LABEL_44": 44,
109
- "LABEL_45": 45,
110
- "LABEL_46": 46,
111
- "LABEL_47": 47,
112
- "LABEL_48": 48,
113
- "LABEL_49": 49,
114
- "LABEL_5": 5,
115
- "LABEL_50": 50,
116
- "LABEL_51": 51,
117
- "LABEL_52": 52,
118
- "LABEL_6": 6,
119
- "LABEL_7": 7,
120
- "LABEL_8": 8,
121
- "LABEL_9": 9
122
- },
123
- "layer_norm_eps": 1e-12,
124
- "max_length": 1024,
125
- "model_type": "audio-spectrogram-transformer",
126
- "num_attention_heads": 12,
127
- "num_hidden_layers": 12,
128
- "num_mel_bins": 128,
129
- "patch_size": 16,
130
- "problem_type": "single_label_classification",
131
- "qkv_bias": true,
132
- "time_stride": 10,
133
- "torch_dtype": "float32",
134
- "transformers_version": "4.40.1"
135
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "MIT/ast-finetuned-audioset-10-10-0.4593",
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": "LABEL_0",
13
+ "1": "LABEL_1",
14
+ "2": "LABEL_2",
15
+ "3": "LABEL_3",
16
+ "4": "LABEL_4",
17
+ "5": "LABEL_5",
18
+ "6": "LABEL_6",
19
+ "7": "LABEL_7",
20
+ "8": "LABEL_8",
21
+ "9": "LABEL_9",
22
+ "10": "LABEL_10",
23
+ "11": "LABEL_11",
24
+ "12": "LABEL_12",
25
+ "13": "LABEL_13",
26
+ "14": "LABEL_14",
27
+ "15": "LABEL_15",
28
+ "16": "LABEL_16",
29
+ "17": "LABEL_17",
30
+ "18": "LABEL_18",
31
+ "19": "LABEL_19",
32
+ "20": "LABEL_20",
33
+ "21": "LABEL_21",
34
+ "22": "LABEL_22",
35
+ "23": "LABEL_23",
36
+ "24": "LABEL_24",
37
+ "25": "LABEL_25",
38
+ "26": "LABEL_26",
39
+ "27": "LABEL_27",
40
+ "28": "LABEL_28",
41
+ "29": "LABEL_29",
42
+ "30": "LABEL_30",
43
+ "31": "LABEL_31",
44
+ "32": "LABEL_32",
45
+ "33": "LABEL_33",
46
+ "34": "LABEL_34",
47
+ "35": "LABEL_35",
48
+ "36": "LABEL_36",
49
+ "37": "LABEL_37",
50
+ "38": "LABEL_38",
51
+ "39": "LABEL_39",
52
+ "40": "LABEL_40",
53
+ "41": "LABEL_41",
54
+ "42": "LABEL_42",
55
+ "43": "LABEL_43",
56
+ "44": "LABEL_44",
57
+ "45": "LABEL_45",
58
+ "46": "LABEL_46",
59
+ "47": "LABEL_47",
60
+ "48": "LABEL_48",
61
+ "49": "LABEL_49",
62
+ "50": "LABEL_50",
63
+ "51": "LABEL_51",
64
+ "52": "LABEL_52",
65
+ "53": "LABEL_53",
66
+ "54": "LABEL_54",
67
+ "55": "LABEL_55",
68
+ "56": "LABEL_56",
69
+ "57": "LABEL_57",
70
+ "58": "LABEL_58",
71
+ "59": "LABEL_59",
72
+ "60": "LABEL_60",
73
+ "61": "LABEL_61",
74
+ "62": "LABEL_62",
75
+ "63": "LABEL_63",
76
+ "64": "LABEL_64",
77
+ "65": "LABEL_65",
78
+ "66": "LABEL_66",
79
+ "67": "LABEL_67",
80
+ "68": "LABEL_68",
81
+ "69": "LABEL_69",
82
+ "70": "LABEL_70",
83
+ "71": "LABEL_71",
84
+ "72": "LABEL_72",
85
+ "73": "LABEL_73",
86
+ "74": "LABEL_74",
87
+ "75": "LABEL_75",
88
+ "76": "LABEL_76",
89
+ "77": "LABEL_77",
90
+ "78": "LABEL_78",
91
+ "79": "LABEL_79",
92
+ "80": "LABEL_80",
93
+ "81": "LABEL_81",
94
+ "82": "LABEL_82",
95
+ "83": "LABEL_83",
96
+ "84": "LABEL_84",
97
+ "85": "LABEL_85"
98
+ },
99
+ "initializer_range": 0.02,
100
+ "intermediate_size": 3072,
101
+ "label2id": {
102
+ "LABEL_0": 0,
103
+ "LABEL_1": 1,
104
+ "LABEL_10": 10,
105
+ "LABEL_11": 11,
106
+ "LABEL_12": 12,
107
+ "LABEL_13": 13,
108
+ "LABEL_14": 14,
109
+ "LABEL_15": 15,
110
+ "LABEL_16": 16,
111
+ "LABEL_17": 17,
112
+ "LABEL_18": 18,
113
+ "LABEL_19": 19,
114
+ "LABEL_2": 2,
115
+ "LABEL_20": 20,
116
+ "LABEL_21": 21,
117
+ "LABEL_22": 22,
118
+ "LABEL_23": 23,
119
+ "LABEL_24": 24,
120
+ "LABEL_25": 25,
121
+ "LABEL_26": 26,
122
+ "LABEL_27": 27,
123
+ "LABEL_28": 28,
124
+ "LABEL_29": 29,
125
+ "LABEL_3": 3,
126
+ "LABEL_30": 30,
127
+ "LABEL_31": 31,
128
+ "LABEL_32": 32,
129
+ "LABEL_33": 33,
130
+ "LABEL_34": 34,
131
+ "LABEL_35": 35,
132
+ "LABEL_36": 36,
133
+ "LABEL_37": 37,
134
+ "LABEL_38": 38,
135
+ "LABEL_39": 39,
136
+ "LABEL_4": 4,
137
+ "LABEL_40": 40,
138
+ "LABEL_41": 41,
139
+ "LABEL_42": 42,
140
+ "LABEL_43": 43,
141
+ "LABEL_44": 44,
142
+ "LABEL_45": 45,
143
+ "LABEL_46": 46,
144
+ "LABEL_47": 47,
145
+ "LABEL_48": 48,
146
+ "LABEL_49": 49,
147
+ "LABEL_5": 5,
148
+ "LABEL_50": 50,
149
+ "LABEL_51": 51,
150
+ "LABEL_52": 52,
151
+ "LABEL_53": 53,
152
+ "LABEL_54": 54,
153
+ "LABEL_55": 55,
154
+ "LABEL_56": 56,
155
+ "LABEL_57": 57,
156
+ "LABEL_58": 58,
157
+ "LABEL_59": 59,
158
+ "LABEL_6": 6,
159
+ "LABEL_60": 60,
160
+ "LABEL_61": 61,
161
+ "LABEL_62": 62,
162
+ "LABEL_63": 63,
163
+ "LABEL_64": 64,
164
+ "LABEL_65": 65,
165
+ "LABEL_66": 66,
166
+ "LABEL_67": 67,
167
+ "LABEL_68": 68,
168
+ "LABEL_69": 69,
169
+ "LABEL_7": 7,
170
+ "LABEL_70": 70,
171
+ "LABEL_71": 71,
172
+ "LABEL_72": 72,
173
+ "LABEL_73": 73,
174
+ "LABEL_74": 74,
175
+ "LABEL_75": 75,
176
+ "LABEL_76": 76,
177
+ "LABEL_77": 77,
178
+ "LABEL_78": 78,
179
+ "LABEL_79": 79,
180
+ "LABEL_8": 8,
181
+ "LABEL_80": 80,
182
+ "LABEL_81": 81,
183
+ "LABEL_82": 82,
184
+ "LABEL_83": 83,
185
+ "LABEL_84": 84,
186
+ "LABEL_85": 85,
187
+ "LABEL_9": 9
188
+ },
189
+ "layer_norm_eps": 1e-12,
190
+ "max_length": 1024,
191
+ "model_type": "audio-spectrogram-transformer",
192
+ "num_attention_heads": 12,
193
+ "num_hidden_layers": 12,
194
+ "num_mel_bins": 128,
195
+ "patch_size": 16,
196
+ "problem_type": "single_label_classification",
197
+ "qkv_bias": true,
198
+ "time_stride": 10,
199
+ "torch_dtype": "float32",
200
+ "transformers_version": "4.40.1"
201
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f6155c7be8ffa923b337332a544dfd02bbb9dc98e52c6f6e1673f24581e1664c
3
- size 344946924
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0dc469732c7a04523e2113de8294c9ec71727f1f453f4111497a690b4ef6853
3
+ size 345048432
preprocessor_config.json CHANGED
@@ -1,13 +1,13 @@
1
- {
2
- "do_normalize": true,
3
- "feature_extractor_type": "ASTFeatureExtractor",
4
- "feature_size": 1,
5
- "max_length": 1024,
6
- "mean": -4.2677393,
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": 4.5689974
13
- }
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "ASTFeatureExtractor",
4
+ "feature_size": 1,
5
+ "max_length": 1024,
6
+ "mean": -4.2677393,
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": 4.5689974
13
+ }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6f35ecf1e0dfa0682b3b421ef261e71c2cd71689839eaf38027099b5a2d1cc2b
3
  size 5048
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad8e16f8cff09268238115046e5f4e9533855ee4ff38343341c3bf344a960241
3
  size 5048