Hugging Face's logo

adrianSauer
/
wav2vec2-wer

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Guaraní
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
adrianSauer commited on
Commit
7a7a24c
•
1 Parent(s): 6487a8a

End of training

Browse files
Files changed (1) hide show
  1. README.md +14 -14
README.md CHANGED
@@ -24,7 +24,7 @@ model-index:
24
  metrics:
25
  - name: Wer
26
  type: wer
27
- value: 43.15937940761636
28
  ---
29
 
30
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -34,8 +34,8 @@ should probably proofread and complete it, then remove this comment. -->
34
 
35
  This model is a fine-tuned version of [glob-asr/wav2vec2-large-xls-r-300m-guarani-small](https://huggingface.co/glob-asr/wav2vec2-large-xls-r-300m-guarani-small) on the Common Voice 16 dataset.
36
  It achieves the following results on the evaluation set:
37
- - Loss: 0.3185
38
- - Wer: 43.1594
39
 
40
  ## Model description
41
 
@@ -60,24 +60,24 @@ The following hyperparameters were used during training:
60
  - seed: 42
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: constant_with_warmup
63
- - lr_scheduler_warmup_steps: 1
64
  - training_steps: 500
65
  - mixed_precision_training: Native AMP
66
 
67
  ### Training results
68
 
69
- | Training Loss | Epoch | Step | Validation Loss | Wer |
70
- |:-------------:|:-----:|:----:|:---------------:|:-------:|
71
- | 0.3978 | 1.01 | 100 | 0.3420 | 46.9111 |
72
- | 0.3263 | 2.02 | 200 | 0.3318 | 46.0931 |
73
- | 0.28 | 3.03 | 300 | 0.3226 | 44.6827 |
74
- | 0.2505 | 4.04 | 400 | 0.3173 | 43.9210 |
75
- | 0.2427 | 5.05 | 500 | 0.3185 | 43.1594 |
76
 
77
 
78
  ### Framework versions
79
 
80
- - Transformers 4.38.2
81
  - Pytorch 2.2.1+cu121
82
- - Datasets 2.18.0
83
- - Tokenizers 0.15.2
 
24
  metrics:
25
  - name: Wer
26
  type: wer
27
+ value: 43.723554301833566
28
  ---
29
 
30
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
34
 
35
  This model is a fine-tuned version of [glob-asr/wav2vec2-large-xls-r-300m-guarani-small](https://huggingface.co/glob-asr/wav2vec2-large-xls-r-300m-guarani-small) on the Common Voice 16 dataset.
36
  It achieves the following results on the evaluation set:
37
+ - Loss: 0.3202
38
+ - Wer: 43.7236
39
 
40
  ## Model description
41
 
 
60
  - seed: 42
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: constant_with_warmup
63
+ - lr_scheduler_warmup_steps: 50
64
  - training_steps: 500
65
  - mixed_precision_training: Native AMP
66
 
67
  ### Training results
68
 
69
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
70
+ |:-------------:|:------:|:----:|:---------------:|:-------:|
71
+ | 0.4174 | 1.0101 | 100 | 0.3535 | 47.6728 |
72
+ | 0.3411 | 2.0202 | 200 | 0.3387 | 46.3188 |
73
+ | 0.2905 | 3.0303 | 300 | 0.3278 | 45.7546 |
74
+ | 0.2591 | 4.0404 | 400 | 0.3214 | 44.6544 |
75
+ | 0.251 | 5.0505 | 500 | 0.3202 | 43.7236 |
76
 
77
 
78
  ### Framework versions
79
 
80
+ - Transformers 4.40.1
81
  - Pytorch 2.2.1+cu121
82
+ - Datasets 2.19.0
83
+ - Tokenizers 0.19.1