sanchit-gandhi HF staff commited on
Commit
3c34bee
1 Parent(s): 565ec30

Training in progress, step 1000

Browse files
config.json ADDED
@@ -0,0 +1,163 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/mms-lid-126",
3
+ "activation_dropout": 0.0,
4
+ "adapter_attn_dim": 16,
5
+ "adapter_kernel_size": 3,
6
+ "adapter_stride": 2,
7
+ "add_adapter": false,
8
+ "apply_spec_augment": true,
9
+ "architectures": [
10
+ "Wav2Vec2ForSequenceClassification"
11
+ ],
12
+ "attention_dropout": 0.1,
13
+ "bos_token_id": 1,
14
+ "classifier_proj_size": 1024,
15
+ "codevector_dim": 1024,
16
+ "contrastive_logits_temperature": 0.1,
17
+ "conv_bias": true,
18
+ "conv_dim": [
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512,
25
+ 512
26
+ ],
27
+ "conv_kernel": [
28
+ 10,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 3,
33
+ 2,
34
+ 2
35
+ ],
36
+ "conv_stride": [
37
+ 5,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2,
43
+ 2
44
+ ],
45
+ "ctc_loss_reduction": "mean",
46
+ "ctc_zero_infinity": false,
47
+ "diversity_loss_weight": 0.1,
48
+ "do_stable_layer_norm": true,
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.1,
56
+ "finetuning_task": "audio-classification",
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.0,
59
+ "hidden_size": 1280,
60
+ "id2label": {
61
+ "0": "Hungarian",
62
+ "1": "Czech",
63
+ "10": "Spanish",
64
+ "11": "Dutch",
65
+ "12": "German",
66
+ "13": "South african",
67
+ "14": "Estonian",
68
+ "15": "Eastern european",
69
+ "16": "Finnish",
70
+ "17": "African",
71
+ "18": "Scottish",
72
+ "19": "Polish",
73
+ "2": "English",
74
+ "20": "Chinese",
75
+ "21": "Indian",
76
+ "22": "Latin american",
77
+ "23": "Lithuanian",
78
+ "24": "Egyptian",
79
+ "3": "Italian",
80
+ "4": "Irish",
81
+ "5": "Vietnamese",
82
+ "6": "French",
83
+ "7": "Romanian",
84
+ "8": "Slovak",
85
+ "9": "American"
86
+ },
87
+ "initializer_range": 0.02,
88
+ "intermediate_size": 5120,
89
+ "label2id": {
90
+ "African": "17",
91
+ "American": "9",
92
+ "Chinese": "20",
93
+ "Czech": "1",
94
+ "Dutch": "11",
95
+ "Eastern european": "15",
96
+ "Egyptian": "24",
97
+ "English": "2",
98
+ "Estonian": "14",
99
+ "Finnish": "16",
100
+ "French": "6",
101
+ "German": "12",
102
+ "Hungarian": "0",
103
+ "Indian": "21",
104
+ "Irish": "4",
105
+ "Italian": "3",
106
+ "Latin american": "22",
107
+ "Lithuanian": "23",
108
+ "Polish": "19",
109
+ "Romanian": "7",
110
+ "Scottish": "18",
111
+ "Slovak": "8",
112
+ "South african": "13",
113
+ "Spanish": "10",
114
+ "Vietnamese": "5"
115
+ },
116
+ "layer_norm_eps": 1e-05,
117
+ "layerdrop": 0.0,
118
+ "mask_feature_length": 10,
119
+ "mask_feature_min_masks": 0,
120
+ "mask_feature_prob": 0.1,
121
+ "mask_time_length": 10,
122
+ "mask_time_min_masks": 2,
123
+ "mask_time_prob": 0.1,
124
+ "model_type": "wav2vec2",
125
+ "num_adapter_layers": 3,
126
+ "num_attention_heads": 16,
127
+ "num_codevector_groups": 2,
128
+ "num_codevectors_per_group": 320,
129
+ "num_conv_pos_embedding_groups": 16,
130
+ "num_conv_pos_embeddings": 128,
131
+ "num_feat_extract_layers": 7,
132
+ "num_hidden_layers": 48,
133
+ "num_negatives": 100,
134
+ "output_hidden_size": 1280,
135
+ "pad_token_id": 0,
136
+ "proj_codevector_dim": 1024,
137
+ "tdnn_dilation": [
138
+ 1,
139
+ 2,
140
+ 3,
141
+ 1,
142
+ 1
143
+ ],
144
+ "tdnn_dim": [
145
+ 512,
146
+ 512,
147
+ 512,
148
+ 512,
149
+ 1500
150
+ ],
151
+ "tdnn_kernel": [
152
+ 5,
153
+ 3,
154
+ 3,
155
+ 1,
156
+ 1
157
+ ],
158
+ "torch_dtype": "float32",
159
+ "transformers_version": "4.38.0.dev0",
160
+ "use_weighted_layer_sum": false,
161
+ "vocab_size": 154,
162
+ "xvector_output_dim": 512
163
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43bc1290d03f49a235eee306a9ac1bebc0d52484f1f2889e2b8a856a3184d06a
3
+ size 3864081756
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,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
runs/Feb16_19-15-14_hf-dgx-01/events.out.tfevents.1708107337.hf-dgx-01.3660307.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b559134414ab1b6e10cee4a9cb2d7ecf03ac308ebf0d8b8d9e87cdcaab46dcb
3
+ size 23188
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abbe5e9d4e0b0f02edf7dcf0ba6b1b6a44761dd5d5840f39cf055bcfd3085f58
3
+ size 4920