Hugging Face's logo

Eimhin03
/
Prereained_Synthetic_Data_augmentation_fine_tuned

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
whisper
Generated from Trainer
Model card Files
xet
 Metrics Community
Eimhin03 commited on
Commit
fe2ec97
·
verified ·
1 Parent(s): 2c77531

Eimhin03/whisper-irish-combined-aug

Browse files
Files changed (2) hide show
  1. README.md +17 -9
  2. model.safetensors +1 -1
README.md CHANGED
@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
18
 
19
  This model is a fine-tuned version of [Eimhin03/output_model_Eubookshop_data_base_model_100k](https://huggingface.co/Eimhin03/output_model_Eubookshop_data_base_model_100k) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 0.8616
22
- - Wer: 41.1890
23
 
24
  ## Model description
25
 
@@ -47,17 +47,25 @@ The following hyperparameters were used during training:
47
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
48
  - lr_scheduler_type: linear
49
  - lr_scheduler_warmup_steps: 1000
50
- - training_steps: 10000
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
- | Training Loss | Epoch | Step | Validation Loss | Wer |
56
- |:-------------:|:-------:|:-----:|:---------------:|:-------:|
57
- | 0.1914 | 8.8347 | 2500 | 0.9888 | 48.0846 |
58
- | 0.0616 | 17.6684 | 5000 | 0.9380 | 47.5215 |
59
- | 0.0211 | 26.5022 | 7500 | 0.8980 | 42.9244 |
60
- | 0.0258 | 35.3360 | 10000 | 0.8616 | 41.1890 |
 
 
 
 
 
 
 
 
61
 
62
 
63
  ### Framework versions
 
18
 
19
  This model is a fine-tuned version of [Eimhin03/output_model_Eubookshop_data_base_model_100k](https://huggingface.co/Eimhin03/output_model_Eubookshop_data_base_model_100k) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 0.8665
22
+ - Wer: 38.6966
23
 
24
  ## Model description
25
 
 
47
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
48
  - lr_scheduler_type: linear
49
  - lr_scheduler_warmup_steps: 1000
50
+ - training_steps: 30000
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
56
+ |:-------------:|:--------:|:-----:|:---------------:|:-------:|
57
+ | 0.1914 | 8.8347 | 2500 | 0.9888 | 48.0846 |
58
+ | 0.0616 | 17.6684 | 5000 | 0.9380 | 47.5215 |
59
+ | 0.0211 | 26.5022 | 7500 | 0.8980 | 42.9244 |
60
+ | 0.0258 | 35.3360 | 10000 | 0.8616 | 41.1890 |
61
+ | 0.0464 | 44.1698 | 12500 | 0.9682 | 44.8352 |
62
+ | 0.0380 | 53.0035 | 15000 | 0.9432 | 43.5429 |
63
+ | 0.0378 | 61.8382 | 17500 | 0.9383 | 41.6782 |
64
+ | 0.0124 | 70.6720 | 20000 | 0.9222 | 41.5859 |
65
+ | 0.0242 | 79.5057 | 22500 | 0.9148 | 40.9582 |
66
+ | 0.0089 | 88.3395 | 25000 | 0.8964 | 42.0474 |
67
+ | 0.0091 | 97.1733 | 27500 | 0.8680 | 38.6966 |
68
+ | 0.0132 | 106.0071 | 30000 | 0.8665 | 38.6966 |
69
 
70
 
71
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:514af8d14a6adead2472a515b164ea9b06a096230435d0152c519d5429a165af
3
  size 290403936
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a9cd8d6902c5f62b5808616ae4e3e21ccff5bd2d3344832b1d0be195d0cf94b
3
  size 290403936