Hugging Face's logo

apoth
/
bird_sounds_classification_2

Audio Classification
Transformers
PyTorch
wav2vec2
Inference Endpoints
Model card Files Community
apoth commited on
Commit
6c56b52
•
1 Parent(s): 7678b89

Upload folder using huggingface_hub

Browse files
Files changed (12) hide show
  1. checkpoint-52500/config.json +205 -0
  2. checkpoint-52500/optimizer.pt +3 -0
  3. checkpoint-52500/preprocessor_config.json +9 -0
  4. checkpoint-52500/pytorch_model.bin +3 -0
  5. checkpoint-52500/rng_state.pth +3 -0
  6. checkpoint-52500/scheduler.pt +3 -0
  7. checkpoint-52500/trainer_state.json +0 -0
  8. checkpoint-52500/training_args.bin +3 -0
  9. config.json +205 -0
  10. preprocessor_config.json +9 -0
  11. pytorch_model.bin +3 -0
  12. training_args.bin +3 -0
checkpoint-52500/config.json ADDED
@@ -0,0 +1,205 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-base-960h",
3
+ "activation_dropout": 0.1,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForSequenceClassification"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 256,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": false,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "sum",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": false,
48
+ "eos_token_id": 2,
49
+ "feat_extract_activation": "gelu",
50
+ "feat_extract_dropout": 0.0,
51
+ "feat_extract_norm": "group",
52
+ "feat_proj_dropout": 0.1,
53
+ "feat_quantizer_dropout": 0.0,
54
+ "final_dropout": 0.1,
55
+ "gradient_checkpointing": false,
56
+ "hidden_act": "gelu",
57
+ "hidden_dropout": 0.1,
58
+ "hidden_dropout_prob": 0.1,
59
+ "hidden_size": 768,
60
+ "id2label": {
61
+ "0": "Andean Guan",
62
+ "1": "Andean Tinamou",
63
+ "2": "Band-tailed Guan",
64
+ "3": "Bartletts Tinamou",
65
+ "4": "Black-capped Tinamou",
66
+ "5": "Blue-throated Piping Guan",
67
+ "6": "Brazilian Tinamou",
68
+ "7": "Brown Tinamou",
69
+ "8": "Cauca Guan",
70
+ "9": "Chaco Chachalaca",
71
+ "10": "Chestnut-winged Chachalaca",
72
+ "11": "Cinereous Tinamou",
73
+ "12": "Colombian Chachalaca",
74
+ "13": "Crested Guan",
75
+ "14": "Dusky Megapode",
76
+ "15": "Dusky-legged Guan",
77
+ "16": "Dwarf Tinamou",
78
+ "17": "Great Tinamou",
79
+ "18": "Grey Tinamou",
80
+ "19": "Grey-headed Chachalaca",
81
+ "20": "Highland Tinamou",
82
+ "21": "Little Chachalaca",
83
+ "22": "Little Tinamou",
84
+ "23": "Orange-footed Scrubfowl",
85
+ "24": "Pale-browed Tinamou",
86
+ "25": "Plain Chachalaca",
87
+ "26": "Red-legged Tinamou",
88
+ "27": "Red-winged Tinamou",
89
+ "28": "Rufous-bellied Chachalaca",
90
+ "29": "Rufous-headed Chachalaca",
91
+ "30": "Rufous-vented Chachalaca",
92
+ "31": "Rusty-margined Guan",
93
+ "32": "Small-billed Tinamou",
94
+ "33": "Solitary Tinamou",
95
+ "34": "Speckled Chachalaca",
96
+ "35": "Spixs Guan",
97
+ "36": "Spotted Nothura",
98
+ "37": "Tataupa Tinamou",
99
+ "38": "Tawny-breasted Tinamou",
100
+ "39": "Thicket Tinamou",
101
+ "40": "Undulated Tinamou",
102
+ "41": "Variegated Tinamou",
103
+ "42": "West Mexican Chachalaca",
104
+ "43": "White-bellied Nothura",
105
+ "44": "White-throated Tinamou",
106
+ "45": "Yellow-legged Tinamou"
107
+ },
108
+ "initializer_range": 0.02,
109
+ "intermediate_size": 3072,
110
+ "label2id": {
111
+ "LABEL_0": 0,
112
+ "LABEL_1": 1,
113
+ "LABEL_10": 10,
114
+ "LABEL_11": 11,
115
+ "LABEL_12": 12,
116
+ "LABEL_13": 13,
117
+ "LABEL_14": 14,
118
+ "LABEL_15": 15,
119
+ "LABEL_16": 16,
120
+ "LABEL_17": 17,
121
+ "LABEL_18": 18,
122
+ "LABEL_19": 19,
123
+ "LABEL_2": 2,
124
+ "LABEL_20": 20,
125
+ "LABEL_21": 21,
126
+ "LABEL_22": 22,
127
+ "LABEL_23": 23,
128
+ "LABEL_24": 24,
129
+ "LABEL_25": 25,
130
+ "LABEL_26": 26,
131
+ "LABEL_27": 27,
132
+ "LABEL_28": 28,
133
+ "LABEL_29": 29,
134
+ "LABEL_3": 3,
135
+ "LABEL_30": 30,
136
+ "LABEL_31": 31,
137
+ "LABEL_32": 32,
138
+ "LABEL_33": 33,
139
+ "LABEL_34": 34,
140
+ "LABEL_35": 35,
141
+ "LABEL_36": 36,
142
+ "LABEL_37": 37,
143
+ "LABEL_38": 38,
144
+ "LABEL_39": 39,
145
+ "LABEL_4": 4,
146
+ "LABEL_40": 40,
147
+ "LABEL_41": 41,
148
+ "LABEL_42": 42,
149
+ "LABEL_43": 43,
150
+ "LABEL_44": 44,
151
+ "LABEL_45": 45,
152
+ "LABEL_5": 5,
153
+ "LABEL_6": 6,
154
+ "LABEL_7": 7,
155
+ "LABEL_8": 8,
156
+ "LABEL_9": 9
157
+ },
158
+ "layer_norm_eps": 1e-05,
159
+ "layerdrop": 0.1,
160
+ "mask_feature_length": 10,
161
+ "mask_feature_min_masks": 0,
162
+ "mask_feature_prob": 0.0,
163
+ "mask_time_length": 10,
164
+ "mask_time_min_masks": 2,
165
+ "mask_time_prob": 0.05,
166
+ "model_type": "wav2vec2",
167
+ "num_adapter_layers": 3,
168
+ "num_attention_heads": 12,
169
+ "num_codevector_groups": 2,
170
+ "num_codevectors_per_group": 320,
171
+ "num_conv_pos_embedding_groups": 16,
172
+ "num_conv_pos_embeddings": 128,
173
+ "num_feat_extract_layers": 7,
174
+ "num_hidden_layers": 12,
175
+ "num_negatives": 100,
176
+ "output_hidden_size": 768,
177
+ "pad_token_id": 0,
178
+ "proj_codevector_dim": 256,
179
+ "tdnn_dilation": [
180
+ 1,
181
+ 2,
182
+ 3,
183
+ 1,
184
+ 1
185
+ ],
186
+ "tdnn_dim": [
187
+ 512,
188
+ 512,
189
+ 512,
190
+ 512,
191
+ 1500
192
+ ],
193
+ "tdnn_kernel": [
194
+ 5,
195
+ 3,
196
+ 3,
197
+ 1,
198
+ 1
199
+ ],
200
+ "torch_dtype": "float32",
201
+ "transformers_version": "4.28.1",
202
+ "use_weighted_layer_sum": false,
203
+ "vocab_size": 32,
204
+ "xvector_output_dim": 512
205
+ }
checkpoint-52500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ddd13c01c317b160d427e017a2488d7cd6b9053ddf89de6b1c2cae8a0ec0e4c1
3
+ size 756767845
checkpoint-52500/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "return_attention_mask": false,
8
+ "sampling_rate": 16000
9
+ }
checkpoint-52500/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f56e80762832fd3578f06072e9bbbc11561f5501ff936313359f7c712292a55
3
+ size 378393747
checkpoint-52500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9438a15ba3efde0103987b091ce417128d74d4ba5854eb3fad7b7c531693ffbd
3
+ size 14631
checkpoint-52500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c62dbce8a251e289100aa9b3c5b18d2eb14fcb9a65c7a9fe1784283fd35ca3df
3
+ size 623
checkpoint-52500/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-52500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d9a1c8e5a297a089174cb2f4de7ca498d4e668761ff83dac049d098125833bd
3
+ size 3567
config.json ADDED
@@ -0,0 +1,205 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-base-960h",
3
+ "activation_dropout": 0.1,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForSequenceClassification"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 256,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": false,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "sum",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": false,
48
+ "eos_token_id": 2,
49
+ "feat_extract_activation": "gelu",
50
+ "feat_extract_dropout": 0.0,
51
+ "feat_extract_norm": "group",
52
+ "feat_proj_dropout": 0.1,
53
+ "feat_quantizer_dropout": 0.0,
54
+ "final_dropout": 0.1,
55
+ "gradient_checkpointing": false,
56
+ "hidden_act": "gelu",
57
+ "hidden_dropout": 0.1,
58
+ "hidden_dropout_prob": 0.1,
59
+ "hidden_size": 768,
60
+ "id2label": {
61
+ "0": "Andean Guan",
62
+ "1": "Andean Tinamou",
63
+ "2": "Band-tailed Guan",
64
+ "3": "Bartletts Tinamou",
65
+ "4": "Black-capped Tinamou",
66
+ "5": "Blue-throated Piping Guan",
67
+ "6": "Brazilian Tinamou",
68
+ "7": "Brown Tinamou",
69
+ "8": "Cauca Guan",
70
+ "9": "Chaco Chachalaca",
71
+ "10": "Chestnut-winged Chachalaca",
72
+ "11": "Cinereous Tinamou",
73
+ "12": "Colombian Chachalaca",
74
+ "13": "Crested Guan",
75
+ "14": "Dusky Megapode",
76
+ "15": "Dusky-legged Guan",
77
+ "16": "Dwarf Tinamou",
78
+ "17": "Great Tinamou",
79
+ "18": "Grey Tinamou",
80
+ "19": "Grey-headed Chachalaca",
81
+ "20": "Highland Tinamou",
82
+ "21": "Little Chachalaca",
83
+ "22": "Little Tinamou",
84
+ "23": "Orange-footed Scrubfowl",
85
+ "24": "Pale-browed Tinamou",
86
+ "25": "Plain Chachalaca",
87
+ "26": "Red-legged Tinamou",
88
+ "27": "Red-winged Tinamou",
89
+ "28": "Rufous-bellied Chachalaca",
90
+ "29": "Rufous-headed Chachalaca",
91
+ "30": "Rufous-vented Chachalaca",
92
+ "31": "Rusty-margined Guan",
93
+ "32": "Small-billed Tinamou",
94
+ "33": "Solitary Tinamou",
95
+ "34": "Speckled Chachalaca",
96
+ "35": "Spixs Guan",
97
+ "36": "Spotted Nothura",
98
+ "37": "Tataupa Tinamou",
99
+ "38": "Tawny-breasted Tinamou",
100
+ "39": "Thicket Tinamou",
101
+ "40": "Undulated Tinamou",
102
+ "41": "Variegated Tinamou",
103
+ "42": "West Mexican Chachalaca",
104
+ "43": "White-bellied Nothura",
105
+ "44": "White-throated Tinamou",
106
+ "45": "Yellow-legged Tinamou"
107
+ },
108
+ "initializer_range": 0.02,
109
+ "intermediate_size": 3072,
110
+ "label2id": {
111
+ "LABEL_0": 0,
112
+ "LABEL_1": 1,
113
+ "LABEL_10": 10,
114
+ "LABEL_11": 11,
115
+ "LABEL_12": 12,
116
+ "LABEL_13": 13,
117
+ "LABEL_14": 14,
118
+ "LABEL_15": 15,
119
+ "LABEL_16": 16,
120
+ "LABEL_17": 17,
121
+ "LABEL_18": 18,
122
+ "LABEL_19": 19,
123
+ "LABEL_2": 2,
124
+ "LABEL_20": 20,
125
+ "LABEL_21": 21,
126
+ "LABEL_22": 22,
127
+ "LABEL_23": 23,
128
+ "LABEL_24": 24,
129
+ "LABEL_25": 25,
130
+ "LABEL_26": 26,
131
+ "LABEL_27": 27,
132
+ "LABEL_28": 28,
133
+ "LABEL_29": 29,
134
+ "LABEL_3": 3,
135
+ "LABEL_30": 30,
136
+ "LABEL_31": 31,
137
+ "LABEL_32": 32,
138
+ "LABEL_33": 33,
139
+ "LABEL_34": 34,
140
+ "LABEL_35": 35,
141
+ "LABEL_36": 36,
142
+ "LABEL_37": 37,
143
+ "LABEL_38": 38,
144
+ "LABEL_39": 39,
145
+ "LABEL_4": 4,
146
+ "LABEL_40": 40,
147
+ "LABEL_41": 41,
148
+ "LABEL_42": 42,
149
+ "LABEL_43": 43,
150
+ "LABEL_44": 44,
151
+ "LABEL_45": 45,
152
+ "LABEL_5": 5,
153
+ "LABEL_6": 6,
154
+ "LABEL_7": 7,
155
+ "LABEL_8": 8,
156
+ "LABEL_9": 9
157
+ },
158
+ "layer_norm_eps": 1e-05,
159
+ "layerdrop": 0.1,
160
+ "mask_feature_length": 10,
161
+ "mask_feature_min_masks": 0,
162
+ "mask_feature_prob": 0.0,
163
+ "mask_time_length": 10,
164
+ "mask_time_min_masks": 2,
165
+ "mask_time_prob": 0.05,
166
+ "model_type": "wav2vec2",
167
+ "num_adapter_layers": 3,
168
+ "num_attention_heads": 12,
169
+ "num_codevector_groups": 2,
170
+ "num_codevectors_per_group": 320,
171
+ "num_conv_pos_embedding_groups": 16,
172
+ "num_conv_pos_embeddings": 128,
173
+ "num_feat_extract_layers": 7,
174
+ "num_hidden_layers": 12,
175
+ "num_negatives": 100,
176
+ "output_hidden_size": 768,
177
+ "pad_token_id": 0,
178
+ "proj_codevector_dim": 256,
179
+ "tdnn_dilation": [
180
+ 1,
181
+ 2,
182
+ 3,
183
+ 1,
184
+ 1
185
+ ],
186
+ "tdnn_dim": [
187
+ 512,
188
+ 512,
189
+ 512,
190
+ 512,
191
+ 1500
192
+ ],
193
+ "tdnn_kernel": [
194
+ 5,
195
+ 3,
196
+ 3,
197
+ 1,
198
+ 1
199
+ ],
200
+ "torch_dtype": "float32",
201
+ "transformers_version": "4.28.1",
202
+ "use_weighted_layer_sum": false,
203
+ "vocab_size": 32,
204
+ "xvector_output_dim": 512
205
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "return_attention_mask": false,
8
+ "sampling_rate": 16000
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f56e80762832fd3578f06072e9bbbc11561f5501ff936313359f7c712292a55
3
+ size 378393747
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d9a1c8e5a297a089174cb2f4de7ca498d4e668761ff83dac049d098125833bd
3
+ size 3567