Hugging Face's logo

pinot
/
wav2vec2-xls-r-1b-ja-phoneme_cv_14_4

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Eval Results
Model card Files Community
1
pinot commited on
Commit
ee3cd8e
·
1 Parent(s): 9f2f071

First model version

Browse files
Files changed (3) hide show
  1. README.md +6 -14
  2. pytorch_model.bin +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: 0.06680282124961409
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 [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the audiofolder dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.2924
36
- - Wer: 0.0668
37
 
38
  ## Model description
39
 
@@ -61,22 +61,14 @@ The following hyperparameters were used during training:
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: linear
63
  - lr_scheduler_warmup_steps: 500
64
- - num_epochs: 1
65
 
66
  ### Training results
67
 
68
  | Training Loss | Epoch | Step | Validation Loss | Wer |
69
  |:-------------:|:-----:|:----:|:---------------:|:------:|
70
- | No log | 0.1 | 400 | 0.8166 | 0.1924 |
71
- | No log | 0.2 | 800 | 0.5377 | 0.1110 |
72
- | No log | 0.29 | 1200 | 0.4711 | 0.1017 |
73
- | No log | 0.39 | 1600 | 0.3988 | 0.0891 |
74
- | 0.9941 | 0.49 | 2000 | 0.3716 | 0.0828 |
75
- | 0.9941 | 0.59 | 2400 | 0.3537 | 0.0784 |
76
- | 0.9941 | 0.69 | 2800 | 0.3339 | 0.0753 |
77
- | 0.9941 | 0.79 | 3200 | 0.3063 | 0.0706 |
78
- | 0.9941 | 0.88 | 3600 | 0.2975 | 0.0692 |
79
- | 0.3454 | 0.98 | 4000 | 0.2924 | 0.0668 |
80
 
81
 
82
  ### Framework versions
 
22
  metrics:
23
  - name: Wer
24
  type: wer
25
+ value: 0.12835735828445224
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 [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the audiofolder dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.5900
36
+ - Wer: 0.1284
37
 
38
  ## Model description
39
 
 
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: linear
63
  - lr_scheduler_warmup_steps: 500
64
+ - num_epochs: 2
65
 
66
  ### Training results
67
 
68
  | Training Loss | Epoch | Step | Validation Loss | Wer |
69
  |:-------------:|:-----:|:----:|:---------------:|:------:|
70
+ | 2.6642 | 0.84 | 400 | 0.8690 | 0.2064 |
71
+ | 0.558 | 1.68 | 800 | 0.5900 | 0.1284 |
 
 
 
 
 
 
 
 
72
 
73
 
74
  ### Framework versions
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:aff2ed23b332170d591a4cd159ec5def1341373d8ce6497ed607173bff926fac
3
  size 3850547117
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f28183416b6e8de0616a76f3451ca3e70d6aa25a6e13cb7403f0dcd22cf3767f
3
  size 3850547117
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:81c0796ca7a3806befdc9701a1867527aace6a5a28fcdc28e2b40c71b6e47bbb
3
  size 4091
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37b6eff97bfb995602902f3ebab77b65fac3d569e6bfaebf37db3db23776467d
3
  size 4091