Hugging Face's logo

tgrhn
/
wav2vec2-turkish-tr-voice

Safetensors
wav2vec2
Generated from Trainer
Model card Files Community
tgrhn commited on
Commit
fd669b1
•
1 Parent(s): 18e0f94

End of training

Browse files
Files changed (5) hide show
  1. README.md +160 -0
  2. config.json +110 -0
  3. model.safetensors +3 -0
  4. preprocessor_config.json +9 -0
  5. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,160 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: facebook/wav2vec2-large
4
+ tags:
5
+ - generated_from_trainer
6
+ model-index:
7
+ - name: wav2vec2-turkish-tr-voice
8
+ results: []
9
+ ---
10
+
11
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
12
+ should probably proofread and complete it, then remove this comment. -->
13
+
14
+ # wav2vec2-turkish-tr-voice
15
+
16
+ This model is a fine-tuned version of [facebook/wav2vec2-large](https://huggingface.co/facebook/wav2vec2-large) on an unknown dataset.
17
+ It achieves the following results on the evaluation set:
18
+ - Loss: 0.0239
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 0.0001
38
+ - train_batch_size: 4
39
+ - eval_batch_size: 4
40
+ - seed: 42
41
+ - distributed_type: multi-GPU
42
+ - num_devices: 2
43
+ - total_train_batch_size: 8
44
+ - total_eval_batch_size: 8
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 1000
48
+ - num_epochs: 50
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss |
54
+ |:-------------:|:------:|:-----:|:---------------:|
55
+ | 2.479 | 0.0765 | 1000 | 0.2672 |
56
+ | 0.3629 | 0.1531 | 2000 | 0.1266 |
57
+ | 0.2485 | 0.2296 | 3000 | 0.1020 |
58
+ | 0.2082 | 0.3062 | 4000 | 0.0916 |
59
+ | 0.1913 | 0.3827 | 5000 | 0.0737 |
60
+ | 0.1757 | 0.4593 | 6000 | 0.0741 |
61
+ | 0.1638 | 0.5358 | 7000 | 0.0686 |
62
+ | 0.1567 | 0.6124 | 8000 | 0.0636 |
63
+ | 0.153 | 0.6889 | 9000 | 0.0608 |
64
+ | 0.1488 | 0.7655 | 10000 | 0.0583 |
65
+ | 0.1374 | 0.8420 | 11000 | 0.0536 |
66
+ | 0.1357 | 0.9186 | 12000 | 0.0511 |
67
+ | 0.1393 | 0.9951 | 13000 | 0.0525 |
68
+ | 0.1314 | 1.0716 | 14000 | 0.0491 |
69
+ | 0.1194 | 1.1482 | 15000 | 0.0498 |
70
+ | 0.1277 | 1.2247 | 16000 | 0.0484 |
71
+ | 0.1173 | 1.3013 | 17000 | 0.0443 |
72
+ | 0.1214 | 1.3778 | 18000 | 0.0443 |
73
+ | 0.1136 | 1.4544 | 19000 | 0.0453 |
74
+ | 0.1107 | 1.5309 | 20000 | 0.0444 |
75
+ | 0.1155 | 1.6075 | 21000 | 0.0419 |
76
+ | 0.1102 | 1.6840 | 22000 | 0.0406 |
77
+ | 0.107 | 1.7606 | 23000 | 0.0414 |
78
+ | 0.104 | 1.8371 | 24000 | 0.0397 |
79
+ | 0.1095 | 1.9137 | 25000 | 0.0405 |
80
+ | 0.1004 | 1.9902 | 26000 | 0.0370 |
81
+ | 0.1038 | 2.0667 | 27000 | 0.0394 |
82
+ | 0.097 | 2.1433 | 28000 | 0.0422 |
83
+ | 0.0983 | 2.2198 | 29000 | 0.0369 |
84
+ | 0.0953 | 2.2964 | 30000 | 0.0405 |
85
+ | 0.0958 | 2.3729 | 31000 | 0.0362 |
86
+ | 0.0934 | 2.4495 | 32000 | 0.0376 |
87
+ | 0.0939 | 2.5260 | 33000 | 0.0348 |
88
+ | 0.0909 | 2.6026 | 34000 | 0.0356 |
89
+ | 0.0927 | 2.6791 | 35000 | 0.0332 |
90
+ | 0.0915 | 2.7557 | 36000 | 0.0324 |
91
+ | 0.0933 | 2.8322 | 37000 | 0.0354 |
92
+ | 0.0905 | 2.9088 | 38000 | 0.0328 |
93
+ | 0.0905 | 2.9853 | 39000 | 0.0351 |
94
+ | 0.0869 | 3.0618 | 40000 | 0.0303 |
95
+ | 0.0869 | 3.1384 | 41000 | 0.0351 |
96
+ | 0.0877 | 3.2149 | 42000 | 0.0339 |
97
+ | 0.0859 | 3.2915 | 43000 | 0.0342 |
98
+ | 0.0922 | 3.3680 | 44000 | 0.0331 |
99
+ | 0.0848 | 3.4446 | 45000 | 0.0350 |
100
+ | 0.0836 | 3.5211 | 46000 | 0.0307 |
101
+ | 0.0887 | 3.5977 | 47000 | 0.0317 |
102
+ | 0.088 | 3.6742 | 48000 | 0.0298 |
103
+ | 0.083 | 3.7508 | 49000 | 0.0325 |
104
+ | 0.0824 | 3.8273 | 50000 | 0.0318 |
105
+ | 0.0811 | 3.9039 | 51000 | 0.0315 |
106
+ | 0.0797 | 3.9804 | 52000 | 0.0289 |
107
+ | 0.0794 | 4.0570 | 53000 | 0.0330 |
108
+ | 0.0844 | 4.1335 | 54000 | 0.0325 |
109
+ | 0.0784 | 4.2100 | 55000 | 0.0330 |
110
+ | 0.0765 | 4.2866 | 56000 | 0.0295 |
111
+ | 0.0793 | 4.3631 | 57000 | 0.0293 |
112
+ | 0.0775 | 4.4397 | 58000 | 0.0302 |
113
+ | 0.076 | 4.5162 | 59000 | 0.0283 |
114
+ | 0.074 | 4.5928 | 60000 | 0.0277 |
115
+ | 0.0766 | 4.6693 | 61000 | 0.0296 |
116
+ | 0.074 | 4.7459 | 62000 | 0.0274 |
117
+ | 0.0744 | 4.8224 | 63000 | 0.0276 |
118
+ | 0.0784 | 4.8990 | 64000 | 0.0295 |
119
+ | 0.0763 | 4.9755 | 65000 | 0.0277 |
120
+ | 0.0722 | 5.0521 | 66000 | 0.0290 |
121
+ | 0.0713 | 5.1286 | 67000 | 0.0277 |
122
+ | 0.0708 | 5.2051 | 68000 | 0.0308 |
123
+ | 0.0714 | 5.2817 | 69000 | 0.0300 |
124
+ | 0.0742 | 5.3582 | 70000 | 0.0273 |
125
+ | 0.0717 | 5.4348 | 71000 | 0.0261 |
126
+ | 0.0699 | 5.5113 | 72000 | 0.0277 |
127
+ | 0.0695 | 5.5879 | 73000 | 0.0275 |
128
+ | 0.0686 | 5.6644 | 74000 | 0.0267 |
129
+ | 0.0702 | 5.7410 | 75000 | 0.0272 |
130
+ | 0.0672 | 5.8175 | 76000 | 0.0269 |
131
+ | 0.0724 | 5.8941 | 77000 | 0.0274 |
132
+ | 0.0696 | 5.9706 | 78000 | 0.0246 |
133
+ | 0.07 | 6.0472 | 79000 | 0.0325 |
134
+ | 0.0903 | 6.1237 | 80000 | 0.0276 |
135
+ | 0.0693 | 6.2002 | 81000 | 0.0250 |
136
+ | 0.0714 | 6.2768 | 82000 | 0.0255 |
137
+ | 0.0655 | 6.3533 | 83000 | 0.0258 |
138
+ | 0.0652 | 6.4299 | 84000 | 0.0270 |
139
+ | 0.0709 | 6.5064 | 85000 | 0.0253 |
140
+ | 0.0666 | 6.5830 | 86000 | 0.0253 |
141
+ | 0.0678 | 6.6595 | 87000 | 0.0257 |
142
+ | 0.0692 | 6.7361 | 88000 | 0.0236 |
143
+ | 0.0657 | 6.8126 | 89000 | 0.0287 |
144
+ | 0.0657 | 6.8892 | 90000 | 0.0240 |
145
+ | 0.0646 | 6.9657 | 91000 | 0.0245 |
146
+ | 0.0616 | 7.0423 | 92000 | 0.0254 |
147
+ | 0.0653 | 7.1188 | 93000 | 0.0291 |
148
+ | 0.0653 | 7.1953 | 94000 | 0.0253 |
149
+ | 0.0617 | 7.2719 | 95000 | 0.0239 |
150
+ | 0.0616 | 7.3484 | 96000 | 0.0245 |
151
+ | 0.0661 | 7.4250 | 97000 | 0.0237 |
152
+ | 0.0629 | 7.5015 | 98000 | 0.0239 |
153
+
154
+
155
+ ### Framework versions
156
+
157
+ - Transformers 4.44.0
158
+ - Pytorch 2.4.0+cu121
159
+ - Datasets 2.18.0
160
+ - Tokenizers 0.19.1
config.json ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-large",
3
+ "activation_dropout": 0.1,
4
+ "adapter_attn_dim": null,
5
+ "adapter_kernel_size": 3,
6
+ "adapter_stride": 2,
7
+ "add_adapter": false,
8
+ "apply_spec_augment": true,
9
+ "architectures": [
10
+ "Wav2Vec2ForCTC"
11
+ ],
12
+ "attention_dropout": 0.1,
13
+ "bos_token_id": 1,
14
+ "classifier_proj_size": 256,
15
+ "codevector_dim": 768,
16
+ "contrastive_logits_temperature": 0.1,
17
+ "conv_bias": false,
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": false,
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "group",
53
+ "feat_proj_dropout": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.1,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_dropout_prob": 0.1,
60
+ "hidden_size": 1024,
61
+ "initializer_range": 0.02,
62
+ "intermediate_size": 4096,
63
+ "layer_norm_eps": 1e-05,
64
+ "layerdrop": 0.1,
65
+ "mask_feature_length": 10,
66
+ "mask_feature_min_masks": 0,
67
+ "mask_feature_prob": 0.0,
68
+ "mask_time_length": 10,
69
+ "mask_time_min_masks": 2,
70
+ "mask_time_prob": 0.05,
71
+ "model_type": "wav2vec2",
72
+ "num_adapter_layers": 3,
73
+ "num_attention_heads": 16,
74
+ "num_codevector_groups": 2,
75
+ "num_codevectors_per_group": 320,
76
+ "num_conv_pos_embedding_groups": 16,
77
+ "num_conv_pos_embeddings": 128,
78
+ "num_feat_extract_layers": 7,
79
+ "num_hidden_layers": 24,
80
+ "num_negatives": 100,
81
+ "output_hidden_size": 1024,
82
+ "pad_token_id": 71,
83
+ "proj_codevector_dim": 768,
84
+ "tdnn_dilation": [
85
+ 1,
86
+ 2,
87
+ 3,
88
+ 1,
89
+ 1
90
+ ],
91
+ "tdnn_dim": [
92
+ 512,
93
+ 512,
94
+ 512,
95
+ 512,
96
+ 1500
97
+ ],
98
+ "tdnn_kernel": [
99
+ 5,
100
+ 3,
101
+ 3,
102
+ 1,
103
+ 1
104
+ ],
105
+ "torch_dtype": "float32",
106
+ "transformers_version": "4.44.0",
107
+ "use_weighted_layer_sum": false,
108
+ "vocab_size": 74,
109
+ "xvector_output_dim": 512
110
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:781b94759c3479873c4bc6ecf5ebca61bf40c1c961f7f68e6f48858cfb0a19dc
3
+ size 1262069568
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": true,
8
+ "sampling_rate": 16000
9
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:510487a731f04a415ff62039dabf3715c1bdf3aeabcf44d374341bebc7376af2
3
+ size 5176