Hugging Face's logo

Solo448
/
SpeechT5-tuned-bn

Text-to-Audio
Transformers
TensorBoard
Safetensors
Bengali
speecht5
Bengali
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
Solo448 commited on
Commit
bdc98b1
•
1 Parent(s): fcdc948

End of training

Browse files
Files changed (3) hide show
  1. README.md +21 -21
  2. model.safetensors +1 -1
  3. runs/Oct16_03-39-35_8fb3c590afc6/events.out.tfevents.1729049978.8fb3c590afc6.810.0 +2 -2
README.md CHANGED
@@ -21,7 +21,7 @@ should probably proofread and complete it, then remove this comment. -->
21
 
22
  This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the train-bn dataset.
23
  It achieves the following results on the evaluation set:
24
- - Loss: 0.6115
25
 
26
  ## Model description
27
 
@@ -40,15 +40,15 @@ More information needed
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
- - learning_rate: 1e-05
44
- - train_batch_size: 16
45
  - eval_batch_size: 2
46
  - seed: 42
47
- - gradient_accumulation_steps: 2
48
  - total_train_batch_size: 32
49
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
50
  - lr_scheduler_type: linear
51
- - lr_scheduler_warmup_steps: 200
52
  - num_epochs: 3.0
53
  - mixed_precision_training: Native AMP
54
 
@@ -56,22 +56,22 @@ The following hyperparameters were used during training:
56
 
57
  | Training Loss | Epoch | Step | Validation Loss |
58
  |:-------------:|:------:|:----:|:---------------:|
59
- | 0.9582 | 0.1805 | 100 | 0.8410 |
60
- | 0.8427 | 0.3610 | 200 | 0.7400 |
61
- | 0.7583 | 0.5415 | 300 | 0.6792 |
62
- | 0.744 | 0.7220 | 400 | 0.6601 |
63
- | 0.7088 | 0.9025 | 500 | 0.6473 |
64
- | 0.6847 | 1.0830 | 600 | 0.6369 |
65
- | 0.6921 | 1.2635 | 700 | 0.6359 |
66
- | 0.7143 | 1.4440 | 800 | 0.6462 |
67
- | 0.6815 | 1.6245 | 900 | 0.6241 |
68
- | 0.6759 | 1.8051 | 1000 | 0.6203 |
69
- | 0.6647 | 1.9856 | 1100 | 0.6171 |
70
- | 0.6558 | 2.1661 | 1200 | 0.6183 |
71
- | 0.6673 | 2.3466 | 1300 | 0.6138 |
72
- | 0.6574 | 2.5271 | 1400 | 0.6135 |
73
- | 0.649 | 2.7076 | 1500 | 0.6125 |
74
- | 0.6546 | 2.8881 | 1600 | 0.6115 |
75
 
76
 
77
  ### Framework versions
 
21
 
22
  This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the train-bn dataset.
23
  It achieves the following results on the evaluation set:
24
+ - Loss: 0.5604
25
 
26
  ## Model description
27
 
 
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
+ - learning_rate: 0.0001
44
+ - train_batch_size: 4
45
  - eval_batch_size: 2
46
  - seed: 42
47
+ - gradient_accumulation_steps: 8
48
  - total_train_batch_size: 32
49
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
50
  - lr_scheduler_type: linear
51
+ - lr_scheduler_warmup_steps: 100
52
  - num_epochs: 3.0
53
  - mixed_precision_training: Native AMP
54
 
 
56
 
57
  | Training Loss | Epoch | Step | Validation Loss |
58
  |:-------------:|:------:|:----:|:---------------:|
59
+ | 0.7779 | 0.1806 | 100 | 0.6979 |
60
+ | 0.6983 | 0.3613 | 200 | 0.6492 |
61
+ | 0.6825 | 0.5419 | 300 | 0.6292 |
62
+ | 0.6571 | 0.7225 | 400 | 0.6200 |
63
+ | 0.6536 | 0.9031 | 500 | 0.6150 |
64
+ | 0.6405 | 1.0838 | 600 | 0.6030 |
65
+ | 0.6319 | 1.2644 | 700 | 0.5925 |
66
+ | 0.6295 | 1.4450 | 800 | 0.5878 |
67
+ | 0.612 | 1.6256 | 900 | 0.5821 |
68
+ | 0.6129 | 1.8063 | 1000 | 0.5774 |
69
+ | 0.6155 | 1.9869 | 1100 | 0.5744 |
70
+ | 0.6056 | 2.1675 | 1200 | 0.5743 |
71
+ | 0.5957 | 2.3482 | 1300 | 0.5735 |
72
+ | 0.5928 | 2.5288 | 1400 | 0.5663 |
73
+ | 0.5974 | 2.7094 | 1500 | 0.5599 |
74
+ | 0.5927 | 2.8900 | 1600 | 0.5604 |
75
 
76
 
77
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b5c205ecfa7837128af0d79136ca9e2cecc2bd2c1a628644df1a3f6fa0fa22d7
3
  size 577789320
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dcad0aa82a3b5203a477eaf39f8dd01d743a3866402ba8eb104dd2ddee2e60c2
3
  size 577789320
runs/Oct16_03-39-35_8fb3c590afc6/events.out.tfevents.1729049978.8fb3c590afc6.810.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:18b3c7563b1501af905924fc99f7235b9622c37e097cb37edb90143b33f93eed
3
- size 24858
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ed27b7f54d7fef42bd8f02b2a63a3dc57c6daef8e034b226b40200ee0dbf3a8
3
+ size 25212