Hugging Face's logo

Samuael
/
ethiopic-asr-characters

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
wav2vec2
Generated from Trainer
Eval Results
Model card Files Metrics Community
Samuael commited on
Commit
3f0c287
·
verified ·
1 Parent(s): aa7dd02

End of training

Browse files
Files changed (3) hide show
  1. README.md +12 -21
  2. model.safetensors +1 -1
  3. training_args.bin +1 -1
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Wer
24
  type: wer
25
- value: 0.28547819213911046
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
32
 
33
  This model is a fine-tuned version of [Samuael/ethiopic-asr-characters](https://huggingface.co/Samuael/ethiopic-asr-characters) on the alffa_amharic dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.3392
36
- - Wer: 0.2855
37
  - Phoneme Cer: 0.0783
38
 
39
  ## Model description
@@ -60,29 +60,20 @@ The following hyperparameters were used during training:
60
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
61
  - lr_scheduler_type: linear
62
  - lr_scheduler_warmup_steps: 100
63
- - num_epochs: 4
64
 
65
  ### Training results
66
 
67
  | Training Loss | Epoch | Step | Validation Loss | Wer | Phoneme Cer |
68
  |:-------------:|:------:|:----:|:---------------:|:------:|:-----------:|
69
- | 0.2465 | 0.2312 | 200 | 0.4203 | 0.3146 | 0.0866 |
70
- | 0.2401 | 0.4624 | 400 | 0.4151 | 0.3057 | 0.0856 |
71
- | 0.2043 | 0.6936 | 600 | 0.3927 | 0.3054 | 0.0837 |
72
- | 0.2115 | 0.9249 | 800 | 0.3945 | 0.3006 | 0.0824 |
73
- | 0.1456 | 1.1561 | 1000 | 0.3864 | 0.2993 | 0.0826 |
74
- | 0.2721 | 1.3873 | 1200 | 0.3873 | 0.3032 | 0.0832 |
75
- | 0.1944 | 1.6185 | 1400 | 0.3796 | 0.2965 | 0.0808 |
76
- | 0.2208 | 1.8497 | 1600 | 0.3637 | 0.2969 | 0.0815 |
77
- | 0.244 | 2.0809 | 1800 | 0.3618 | 0.2940 | 0.0811 |
78
- | 0.265 | 2.3121 | 2000 | 0.3615 | 0.2928 | 0.0810 |
79
- | 0.2747 | 2.5434 | 2200 | 0.3514 | 0.2907 | 0.0794 |
80
- | 0.4672 | 2.7746 | 2400 | 0.3562 | 0.2888 | 0.0793 |
81
- | 0.3279 | 3.0058 | 2600 | 0.3478 | 0.2863 | 0.0780 |
82
- | 0.5459 | 3.2370 | 2800 | 0.3412 | 0.2856 | 0.0788 |
83
- | 0.2769 | 3.4682 | 3000 | 0.3412 | 0.2860 | 0.0783 |
84
- | 0.3112 | 3.6994 | 3200 | 0.3402 | 0.2875 | 0.0789 |
85
- | 0.3468 | 3.9306 | 3400 | 0.3392 | 0.2855 | 0.0783 |
86
 
87
 
88
  ### Framework versions
 
22
  metrics:
23
  - name: Wer
24
  type: wer
25
+ value: 0.2861248832363297
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
32
 
33
  This model is a fine-tuned version of [Samuael/ethiopic-asr-characters](https://huggingface.co/Samuael/ethiopic-asr-characters) on the alffa_amharic dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.3784
36
+ - Wer: 0.2861
37
  - Phoneme Cer: 0.0783
38
 
39
  ## Model description
 
60
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
61
  - lr_scheduler_type: linear
62
  - lr_scheduler_warmup_steps: 100
63
+ - num_epochs: 2
64
 
65
  ### Training results
66
 
67
  | Training Loss | Epoch | Step | Validation Loss | Wer | Phoneme Cer |
68
  |:-------------:|:------:|:----:|:---------------:|:------:|:-----------:|
69
+ | 0.1838 | 0.2312 | 200 | 0.4123 | 0.2914 | 0.0801 |
70
+ | 0.144 | 0.4624 | 400 | 0.4168 | 0.2917 | 0.0792 |
71
+ | 0.1087 | 0.6936 | 600 | 0.4104 | 0.2904 | 0.0785 |
72
+ | 0.1421 | 0.9249 | 800 | 0.3881 | 0.2878 | 0.0793 |
73
+ | 0.105 | 1.1561 | 1000 | 0.3944 | 0.2896 | 0.0781 |
74
+ | 0.189 | 1.3873 | 1200 | 0.3918 | 0.2895 | 0.0793 |
75
+ | 0.1389 | 1.6185 | 1400 | 0.3844 | 0.2886 | 0.0782 |
76
+ | 0.1623 | 1.8497 | 1600 | 0.3784 | 0.2861 | 0.0783 |
 
 
 
 
 
 
 
 
 
77
 
78
 
79
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cd2c99933821df7e05349a293786a3d53687d017c78870b27d931a227c248c12
3
  size 1262791480
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf3ec969a8d8eb72e07eb36cead6187264aa29b461a251615b335db641923bc9
3
  size 1262791480
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d5cb5785a2b15e3aed6d250388d88caa8581e6a6063e15b0be5571c350f93989
3
  size 5240
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f078fd1392d86a13308d4e7c70f151aed9e2dcd36572ae9463e70d290b6f5122
3
  size 5240