Hugging Face's logo

MelanieKoe
/
w2v2-base-pretrained_lr1e-4_at0.8_da0.6

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MelanieKoe commited on
Commit
d3f552f
•
1 Parent(s): f0cc052

Model save

Browse files
Files changed (2) hide show
  1. README.md +20 -62
  2. model.safetensors +1 -1
README.md CHANGED
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 3.3840
21
- - Wer: 1.0
22
 
23
  ## Model description
24
 
@@ -43,70 +43,28 @@ The following hyperparameters were used during training:
43
  - seed: 42
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
  - lr_scheduler_type: linear
46
- - lr_scheduler_warmup_steps: 100
47
- - num_epochs: 200
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
- | Training Loss | Epoch | Step | Validation Loss | Wer |
53
- |:-------------:|:------:|:----:|:---------------:|:---:|
54
- | 17.8105 | 3.57 | 100 | 3.5095 | 1.0 |
55
- | 3.2528 | 7.14 | 200 | 3.5113 | 1.0 |
56
- | 3.2064 | 10.71 | 300 | 3.3465 | 1.0 |
57
- | 3.2083 | 14.29 | 400 | 3.2972 | 1.0 |
58
- | 3.1863 | 17.86 | 500 | 3.4229 | 1.0 |
59
- | 3.1881 | 21.43 | 600 | 3.3365 | 1.0 |
60
- | 3.1875 | 25.0 | 700 | 3.4389 | 1.0 |
61
- | 3.1814 | 28.57 | 800 | 3.3845 | 1.0 |
62
- | 3.184 | 32.14 | 900 | 3.2846 | 1.0 |
63
- | 3.1834 | 35.71 | 1000 | 3.4072 | 1.0 |
64
- | 3.1872 | 39.29 | 1100 | 3.3569 | 1.0 |
65
- | 3.1793 | 42.86 | 1200 | 3.3841 | 1.0 |
66
- | 3.1852 | 46.43 | 1300 | 3.4166 | 1.0 |
67
- | 3.1829 | 50.0 | 1400 | 3.3862 | 1.0 |
68
- | 3.1792 | 53.57 | 1500 | 3.3912 | 1.0 |
69
- | 3.1824 | 57.14 | 1600 | 3.3566 | 1.0 |
70
- | 3.177 | 60.71 | 1700 | 3.3635 | 1.0 |
71
- | 3.1798 | 64.29 | 1800 | 3.4135 | 1.0 |
72
- | 3.179 | 67.86 | 1900 | 3.4062 | 1.0 |
73
- | 3.1797 | 71.43 | 2000 | 3.3989 | 1.0 |
74
- | 3.1807 | 75.0 | 2100 | 3.3615 | 1.0 |
75
- | 3.1763 | 78.57 | 2200 | 3.4078 | 1.0 |
76
- | 3.1835 | 82.14 | 2300 | 3.4158 | 1.0 |
77
- | 3.1796 | 85.71 | 2400 | 3.3696 | 1.0 |
78
- | 3.1798 | 89.29 | 2500 | 3.3899 | 1.0 |
79
- | 3.1755 | 92.86 | 2600 | 3.3690 | 1.0 |
80
- | 3.1783 | 96.43 | 2700 | 3.3968 | 1.0 |
81
- | 3.1805 | 100.0 | 2800 | 3.4242 | 1.0 |
82
- | 3.1772 | 103.57 | 2900 | 3.3782 | 1.0 |
83
- | 3.179 | 107.14 | 3000 | 3.3761 | 1.0 |
84
- | 3.177 | 110.71 | 3100 | 3.3330 | 1.0 |
85
- | 3.1806 | 114.29 | 3200 | 3.3641 | 1.0 |
86
- | 3.1742 | 117.86 | 3300 | 3.3969 | 1.0 |
87
- | 3.1823 | 121.43 | 3400 | 3.4045 | 1.0 |
88
- | 3.1813 | 125.0 | 3500 | 3.4124 | 1.0 |
89
- | 3.1771 | 128.57 | 3600 | 3.3649 | 1.0 |
90
- | 3.1795 | 132.14 | 3700 | 3.3803 | 1.0 |
91
- | 3.1771 | 135.71 | 3800 | 3.3956 | 1.0 |
92
- | 3.1795 | 139.29 | 3900 | 3.3762 | 1.0 |
93
- | 3.1734 | 142.86 | 4000 | 3.3650 | 1.0 |
94
- | 3.1769 | 146.43 | 4100 | 3.3762 | 1.0 |
95
- | 3.1789 | 150.0 | 4200 | 3.3805 | 1.0 |
96
- | 3.1771 | 153.57 | 4300 | 3.3740 | 1.0 |
97
- | 3.1772 | 157.14 | 4400 | 3.3742 | 1.0 |
98
- | 3.177 | 160.71 | 4500 | 3.3985 | 1.0 |
99
- | 3.1785 | 164.29 | 4600 | 3.3759 | 1.0 |
100
- | 3.1749 | 167.86 | 4700 | 3.3883 | 1.0 |
101
- | 3.1763 | 171.43 | 4800 | 3.4118 | 1.0 |
102
- | 3.1798 | 175.0 | 4900 | 3.3929 | 1.0 |
103
- | 3.176 | 178.57 | 5000 | 3.3799 | 1.0 |
104
- | 3.1749 | 182.14 | 5100 | 3.4036 | 1.0 |
105
- | 3.1778 | 185.71 | 5200 | 3.3932 | 1.0 |
106
- | 3.1802 | 189.29 | 5300 | 3.3977 | 1.0 |
107
- | 3.1746 | 192.86 | 5400 | 3.3877 | 1.0 |
108
- | 3.1776 | 196.43 | 5500 | 3.3846 | 1.0 |
109
- | 3.1759 | 200.0 | 5600 | 3.3840 | 1.0 |
110
 
111
 
112
  ### Framework versions
 
17
 
18
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 2.5777
21
+ - Wer: 0.1820
22
 
23
  ## Model description
24
 
 
43
  - seed: 42
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
  - lr_scheduler_type: linear
46
+ - lr_scheduler_warmup_steps: 1000
47
+ - training_steps: 3500
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
53
+ |:-------------:|:------:|:----:|:---------------:|:------:|
54
+ | 15.4716 | 8.93 | 250 | 3.5805 | 1.0 |
55
+ | 3.156 | 17.86 | 500 | 3.1574 | 1.0 |
56
+ | 2.6095 | 26.79 | 750 | 1.2490 | 1.0 |
57
+ | 0.2716 | 35.71 | 1000 | 1.6097 | 0.2422 |
58
+ | 0.0901 | 44.64 | 1250 | 1.8537 | 0.2085 |
59
+ | 0.0526 | 53.57 | 1500 | 1.7249 | 0.2080 |
60
+ | 0.0328 | 62.5 | 1750 | 2.2934 | 0.1922 |
61
+ | 0.0237 | 71.43 | 2000 | 2.1689 | 0.1948 |
62
+ | 0.018 | 80.36 | 2250 | 2.4606 | 0.1918 |
63
+ | 0.0144 | 89.29 | 2500 | 2.6943 | 0.1837 |
64
+ | 0.0108 | 98.21 | 2750 | 2.5514 | 0.1807 |
65
+ | 0.0078 | 107.14 | 3000 | 2.5035 | 0.1858 |
66
+ | 0.0064 | 116.07 | 3250 | 2.6022 | 0.1845 |
67
+ | 0.0069 | 125.0 | 3500 | 2.5777 | 0.1820 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
68
 
69
 
70
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:77ab8295a181aff895643e9865a560364046e8d11247d11f26fc73e9e98f63a4
3
  size 377629528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa494635dbc79230feecbf02c48e38755bf0afa05375a7f0b0f5a70a953fdb49
3
  size 377629528