Hugging Face's logo

GiorgiSekhniashvili
/
whisper-tiny-ka-02

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Georgian
whisper
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
GiorgiSekhniashvili commited on
Commit
a6ca647
•
1 Parent(s): 916c9de

End of training

Browse files
Files changed (2) hide show
  1. README.md +13 -9
  2. runs/Mar04_03-55-52_razer-14/events.out.tfevents.1709510155.razer-14.376.0 +2 -2
README.md CHANGED
@@ -25,7 +25,7 @@ model-index:
25
  metrics:
26
  - name: Wer
27
  type: wer
28
- value: 100.0
29
  ---
30
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -35,8 +35,8 @@ should probably proofread and complete it, then remove this comment. -->
35
 
36
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice dataset.
37
  It achieves the following results on the evaluation set:
38
- - Loss: 7.6994
39
- - Wer: 100.0
40
 
41
  ## Model description
42
 
@@ -56,22 +56,26 @@ More information needed
56
 
57
  The following hyperparameters were used during training:
58
  - learning_rate: 0.003
59
- - train_batch_size: 8
60
  - eval_batch_size: 8
61
  - seed: 42
62
- - gradient_accumulation_steps: 8
63
  - total_train_batch_size: 64
64
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
65
  - lr_scheduler_type: linear
66
  - lr_scheduler_warmup_steps: 1000
67
- - training_steps: 30
68
  - mixed_precision_training: Native AMP
69
 
70
  ### Training results
71
 
72
- | Training Loss | Epoch | Step | Validation Loss | Wer |
73
- |:-------------:|:-----:|:----:|:---------------:|:-----:|
74
- | 8.8508 | 15.38 | 25 | 7.6994 | 100.0 |
 
 
 
 
75
 
76
 
77
  ### Framework versions
 
25
  metrics:
26
  - name: Wer
27
  type: wer
28
+ value: 134.83959527973963
29
  ---
30
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
35
 
36
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice dataset.
37
  It achieves the following results on the evaluation set:
38
+ - Loss: 5.0988
39
+ - Wer: 134.8396
40
 
41
  ## Model description
42
 
 
56
 
57
  The following hyperparameters were used during training:
58
  - learning_rate: 0.003
59
+ - train_batch_size: 16
60
  - eval_batch_size: 8
61
  - seed: 42
62
+ - gradient_accumulation_steps: 4
63
  - total_train_batch_size: 64
64
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
65
  - lr_scheduler_type: linear
66
  - lr_scheduler_warmup_steps: 1000
67
+ - training_steps: 5000
68
  - mixed_precision_training: Native AMP
69
 
70
  ### Training results
71
 
72
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
73
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
74
+ | 5.0576 | 1.45 | 1000 | 5.1841 | 158.1475 |
75
+ | 4.6405 | 2.9 | 2000 | 4.8881 | 131.9237 |
76
+ | 4.0627 | 4.35 | 3000 | 4.9336 | 143.5772 |
77
+ | 3.781 | 5.8 | 4000 | 4.9113 | 129.0976 |
78
+ | 3.0831 | 7.25 | 5000 | 5.0988 | 134.8396 |
79
 
80
 
81
  ### Framework versions
runs/Mar04_03-55-52_razer-14/events.out.tfevents.1709510155.razer-14.376.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c9f7cc72f078f27d22c772c40d808895ea5c772aab34816d8e9ff01c1f808dd1
3
- size 40476
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22f3ca28464a89374e492bbd7a0ee0bd4fd135dacc69561fa2429eaae3f879e4
3
+ size 49588