Hugging Face's logo

Ocelotr
/
speecht5_tts-sil

Text-to-Audio
Transformers
PyTorch
TensorBoard
Arabic
speecht5
ara
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
1
Ocelotr commited on
Commit
4931192
•
1 Parent(s): fc0af9b

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +104 -0
README.md ADDED
@@ -0,0 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - ar
4
+ license: mit
5
+ tags:
6
+ - ara
7
+ - generated_from_trainer
8
+ datasets:
9
+ - SDA_CLEAN_NAJDI
10
+ model-index:
11
+ - name: SpeechT5 TTS
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ # SpeechT5 TTS
19
+
20
+ This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the SDA dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 0.4853
23
+
24
+ ## Model description
25
+
26
+ More information needed
27
+
28
+ ## Intended uses & limitations
29
+
30
+ More information needed
31
+
32
+ ## Training and evaluation data
33
+
34
+ More information needed
35
+
36
+ ## Training procedure
37
+
38
+ ### Training hyperparameters
39
+
40
+ The following hyperparameters were used during training:
41
+ - learning_rate: 1e-05
42
+ - train_batch_size: 16
43
+ - eval_batch_size: 8
44
+ - seed: 42
45
+ - gradient_accumulation_steps: 2
46
+ - total_train_batch_size: 32
47
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
+ - lr_scheduler_type: linear
49
+ - lr_scheduler_warmup_steps: 500
50
+ - training_steps: 40000
51
+ - mixed_precision_training: Native AMP
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:-----:|:-----:|:---------------:|
57
+ | 0.5703 | 1.49 | 1000 | 0.5289 |
58
+ | 0.541 | 2.98 | 2000 | 0.5131 |
59
+ | 0.5487 | 4.46 | 3000 | 0.5059 |
60
+ | 0.5232 | 5.95 | 4000 | 0.5011 |
61
+ | 0.5295 | 7.44 | 5000 | 0.4979 |
62
+ | 0.5257 | 8.93 | 6000 | 0.4970 |
63
+ | 0.5091 | 10.42 | 7000 | 0.4905 |
64
+ | 0.5141 | 11.9 | 8000 | 0.4893 |
65
+ | 0.5033 | 13.39 | 9000 | 0.4865 |
66
+ | 0.507 | 14.88 | 10000 | 0.4850 |
67
+ | 0.502 | 16.37 | 11000 | 0.4830 |
68
+ | 0.497 | 17.86 | 12000 | 0.4823 |
69
+ | 0.4974 | 19.35 | 13000 | 0.4801 |
70
+ | 0.4993 | 20.83 | 14000 | 0.4794 |
71
+ | 0.496 | 22.32 | 15000 | 0.4814 |
72
+ | 0.4845 | 23.81 | 16000 | 0.4780 |
73
+ | 0.4977 | 25.3 | 17000 | 0.4775 |
74
+ | 0.4888 | 26.79 | 18000 | 0.4780 |
75
+ | 0.4773 | 28.27 | 19000 | 0.4792 |
76
+ | 0.4914 | 29.76 | 20000 | 0.4817 |
77
+ | 0.4864 | 31.25 | 21000 | 0.4775 |
78
+ | 0.486 | 32.74 | 22000 | 0.4773 |
79
+ | 0.4884 | 34.23 | 23000 | 0.4835 |
80
+ | 0.4856 | 35.71 | 24000 | 0.4788 |
81
+ | 0.4814 | 37.2 | 25000 | 0.4811 |
82
+ | 0.4831 | 38.69 | 26000 | 0.4814 |
83
+ | 0.4732 | 40.18 | 27000 | 0.4816 |
84
+ | 0.4846 | 41.67 | 28000 | 0.4812 |
85
+ | 0.4731 | 43.15 | 29000 | 0.4843 |
86
+ | 0.4772 | 44.64 | 30000 | 0.4830 |
87
+ | 0.4793 | 46.13 | 31000 | 0.4834 |
88
+ | 0.4736 | 47.62 | 32000 | 0.4834 |
89
+ | 0.4798 | 49.11 | 33000 | 0.4826 |
90
+ | 0.4744 | 50.6 | 34000 | 0.4841 |
91
+ | 0.4784 | 52.08 | 35000 | 0.4844 |
92
+ | 0.4743 | 53.57 | 36000 | 0.4851 |
93
+ | 0.4779 | 55.06 | 37000 | 0.4854 |
94
+ | 0.4719 | 56.55 | 38000 | 0.4854 |
95
+ | 0.4825 | 58.04 | 39000 | 0.4856 |
96
+ | 0.4805 | 59.52 | 40000 | 0.4853 |
97
+
98
+
99
+ ### Framework versions
100
+
101
+ - Transformers 4.30.0.dev0
102
+ - Pytorch 2.0.1+cu117
103
+ - Datasets 2.13.0
104
+ - Tokenizers 0.13.3