Hugging Face's logo

tylee4
/
lr_3e-5

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Korean
whisper
Generated from Trainer
Model card Files
xet
Metrics Community
tylee4 commited on
Commit
63d95cc
·
verified ·
1 Parent(s): 43dcbd7

End of training

Browse files
Files changed (1) hide show
  1. README.md +13 -13
README.md CHANGED
@@ -20,8 +20,8 @@ should probably proofread and complete it, then remove this comment. -->
20
 
21
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the zeroth_korean dataset.
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.1102
24
- - Cer: 3.8315
25
 
26
  ## Model description
27
 
@@ -40,7 +40,7 @@ More information needed
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
- - learning_rate: 3e-05
44
  - train_batch_size: 16
45
  - eval_batch_size: 8
46
  - seed: 42
@@ -54,16 +54,16 @@ The following hyperparameters were used during training:
54
 
55
  | Training Loss | Epoch | Step | Validation Loss | Cer |
56
  |:-------------:|:------:|:----:|:---------------:|:-------:|
57
- | 0.4247 | 0.4488 | 500 | 0.4256 | 14.4414 |
58
- | 0.3054 | 0.8977 | 1000 | 0.2914 | 9.9730 |
59
- | 0.1796 | 1.3465 | 1500 | 0.2246 | 7.9640 |
60
- | 0.1326 | 1.7953 | 2000 | 0.1814 | 6.4452 |
61
- | 0.0678 | 2.2442 | 2500 | 0.1527 | 5.5537 |
62
- | 0.0565 | 2.6930 | 3000 | 0.1364 | 4.8596 |
63
- | 0.0301 | 3.1418 | 3500 | 0.1242 | 4.5292 |
64
- | 0.0296 | 3.5907 | 4000 | 0.1167 | 4.1332 |
65
- | 0.0134 | 4.0395 | 4500 | 0.1113 | 3.9040 |
66
- | 0.0162 | 4.4883 | 5000 | 0.1102 | 3.8315 |
67
 
68
 
69
  ### Framework versions
 
20
 
21
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the zeroth_korean dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.3568
24
+ - Cer: 11.5477
25
 
26
  ## Model description
27
 
 
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
+ - learning_rate: 3e-06
44
  - train_batch_size: 16
45
  - eval_batch_size: 8
46
  - seed: 42
 
54
 
55
  | Training Loss | Epoch | Step | Validation Loss | Cer |
56
  |:-------------:|:------:|:----:|:---------------:|:-------:|
57
+ | 0.5846 | 0.4488 | 500 | 0.5778 | 17.0684 |
58
+ | 0.5078 | 0.8977 | 1000 | 0.4889 | 14.7653 |
59
+ | 0.438 | 1.3465 | 1500 | 0.4454 | 13.8077 |
60
+ | 0.3945 | 1.7953 | 2000 | 0.4158 | 12.9504 |
61
+ | 0.3629 | 2.2442 | 2500 | 0.3952 | 12.4517 |
62
+ | 0.3363 | 2.6930 | 3000 | 0.3804 | 12.3059 |
63
+ | 0.3301 | 3.1418 | 3500 | 0.3699 | 11.8438 |
64
+ | 0.3291 | 3.5907 | 4000 | 0.3626 | 11.7568 |
65
+ | 0.2829 | 4.0395 | 4500 | 0.3582 | 11.5566 |
66
+ | 0.3056 | 4.4883 | 5000 | 0.3568 | 11.5477 |
67
 
68
 
69
  ### Framework versions