Hugging Face's logo

kresnik
/
wav2vec2-large-xls-r-300m-korean

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
kresnik commited on
Commit
7008951
•
1 Parent(s): eb14df2

End of training

Browse files
Files changed (1) hide show
  1. README.md +54 -10
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Wer
24
  type: wer
25
- value: 0.6404833836858006
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-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice_13_0 dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 1.0384
36
- - Wer: 0.6405
37
 
38
  ## Model description
39
 
@@ -61,16 +61,60 @@ 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: 30
65
 
66
  ### Training results
67
 
68
- | Training Loss | Epoch | Step | Validation Loss | Wer |
69
- |:-------------:|:-----:|:----:|:---------------:|:------:|
70
- | 20.1128 | 6.25 | 400 | 4.7129 | 1.0 |
71
- | 4.0984 | 12.5 | 800 | 2.3725 | 0.9255 |
72
- | 1.547 | 18.75 | 1200 | 1.2406 | 0.6959 |
73
- | 0.6862 | 25.0 | 1600 | 1.0384 | 0.6405 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
74
 
75
 
76
  ### Framework versions
 
22
  metrics:
23
  - name: Wer
24
  type: wer
25
+ value: 0.5931520644511581
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-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice_13_0 dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 1.4687
36
+ - Wer: 0.5932
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: 300
65
 
66
  ### Training results
67
 
68
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
69
+ |:-------------:|:------:|:-----:|:---------------:|:------:|
70
+ | 20.8922 | 6.25 | 400 | 4.6827 | 0.9990 |
71
+ | 4.0513 | 12.5 | 800 | 2.3657 | 0.9204 |
72
+ | 1.5386 | 18.75 | 1200 | 1.2355 | 0.7392 |
73
+ | 0.7429 | 25.0 | 1600 | 1.1179 | 0.6636 |
74
+ | 0.3746 | 31.25 | 2000 | 1.0465 | 0.6314 |
75
+ | 0.2407 | 37.5 | 2400 | 1.1492 | 0.6596 |
76
+ | 0.1966 | 43.75 | 2800 | 1.1291 | 0.6344 |
77
+ | 0.1697 | 50.0 | 3200 | 1.1897 | 0.6395 |
78
+ | 0.1533 | 56.25 | 3600 | 1.2202 | 0.6193 |
79
+ | 0.129 | 62.5 | 4000 | 1.2106 | 0.6516 |
80
+ | 0.1097 | 68.75 | 4400 | 1.1662 | 0.6254 |
81
+ | 0.102 | 75.0 | 4800 | 1.2086 | 0.6133 |
82
+ | 0.0918 | 81.25 | 5200 | 1.2295 | 0.6485 |
83
+ | 0.0806 | 87.5 | 5600 | 1.2861 | 0.6123 |
84
+ | 0.0738 | 93.75 | 6000 | 1.2436 | 0.6093 |
85
+ | 0.0697 | 100.0 | 6400 | 1.3496 | 0.6626 |
86
+ | 0.0667 | 106.25 | 6800 | 1.2364 | 0.6133 |
87
+ | 0.0591 | 112.5 | 7200 | 1.2689 | 0.6062 |
88
+ | 0.054 | 118.75 | 7600 | 1.2886 | 0.6183 |
89
+ | 0.0523 | 125.0 | 8000 | 1.3328 | 0.6445 |
90
+ | 0.0542 | 131.25 | 8400 | 1.4019 | 0.6133 |
91
+ | 0.045 | 137.5 | 8800 | 1.3426 | 0.6042 |
92
+ | 0.0425 | 143.75 | 9200 | 1.3042 | 0.6032 |
93
+ | 0.0378 | 150.0 | 9600 | 1.3638 | 0.6224 |
94
+ | 0.0354 | 156.25 | 10000 | 1.3397 | 0.6294 |
95
+ | 0.0282 | 162.5 | 10400 | 1.3939 | 0.6173 |
96
+ | 0.0288 | 168.75 | 10800 | 1.3674 | 0.6475 |
97
+ | 0.0278 | 175.0 | 11200 | 1.3636 | 0.6324 |
98
+ | 0.0239 | 181.25 | 11600 | 1.4101 | 0.6405 |
99
+ | 0.0238 | 187.5 | 12000 | 1.4528 | 0.6163 |
100
+ | 0.0214 | 193.75 | 12400 | 1.4458 | 0.6093 |
101
+ | 0.0194 | 200.0 | 12800 | 1.3920 | 0.6304 |
102
+ | 0.0168 | 206.25 | 13200 | 1.4277 | 0.6193 |
103
+ | 0.0168 | 212.5 | 13600 | 1.3959 | 0.6203 |
104
+ | 0.0154 | 218.75 | 14000 | 1.4043 | 0.6133 |
105
+ | 0.0144 | 225.0 | 14400 | 1.4508 | 0.6193 |
106
+ | 0.0134 | 231.25 | 14800 | 1.4309 | 0.6224 |
107
+ | 0.0109 | 237.5 | 15200 | 1.4301 | 0.6123 |
108
+ | 0.0107 | 243.75 | 15600 | 1.4373 | 0.6002 |
109
+ | 0.0098 | 250.0 | 16000 | 1.4147 | 0.6113 |
110
+ | 0.0095 | 256.25 | 16400 | 1.4585 | 0.6193 |
111
+ | 0.009 | 262.5 | 16800 | 1.4424 | 0.6203 |
112
+ | 0.0079 | 268.75 | 17200 | 1.5019 | 0.6193 |
113
+ | 0.0066 | 275.0 | 17600 | 1.4835 | 0.5932 |
114
+ | 0.0059 | 281.25 | 18000 | 1.4749 | 0.5992 |
115
+ | 0.0057 | 287.5 | 18400 | 1.4897 | 0.6002 |
116
+ | 0.0053 | 293.75 | 18800 | 1.4667 | 0.5901 |
117
+ | 0.0048 | 300.0 | 19200 | 1.4687 | 0.5932 |
118
 
119
 
120
  ### Framework versions