Hugging Face's logo

MelanieKoe
/
w2v2-base-pretrained_lr5e-5_at0.8_da0.5

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MelanieKoe commited on
Commit
aeb60b6
·
verified ·
1 Parent(s): c25a193

Model save

Browse files
Files changed (2) hide show
  1. README.md +19 -19
  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: 1.9030
21
- - Wer: 0.2008
22
 
23
  ## Model description
24
 
@@ -43,7 +43,7 @@ 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: 500
47
  - training_steps: 4000
48
  - mixed_precision_training: Native AMP
49
 
@@ -51,22 +51,22 @@ The following hyperparameters were used during training:
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Wer |
53
  |:-------------:|:------:|:----:|:---------------:|:------:|
54
- | 15.4573 | 10.87 | 250 | 3.6074 | 1.0 |
55
- | 3.1552 | 21.74 | 500 | 3.1637 | 1.0 |
56
- | 3.0394 | 32.61 | 750 | 3.0916 | 1.0 |
57
- | 1.9194 | 43.48 | 1000 | 1.1132 | 0.5109 |
58
- | 0.1754 | 54.35 | 1250 | 1.3050 | 0.2439 |
59
- | 0.0778 | 65.22 | 1500 | 1.5929 | 0.2392 |
60
- | 0.0511 | 76.09 | 1750 | 1.3839 | 0.2213 |
61
- | 0.04 | 86.96 | 2000 | 1.5492 | 0.2179 |
62
- | 0.0303 | 97.83 | 2250 | 1.6937 | 0.1982 |
63
- | 0.0228 | 108.7 | 2500 | 1.7444 | 0.2029 |
64
- | 0.0193 | 119.57 | 2750 | 1.8009 | 0.1982 |
65
- | 0.0157 | 130.43 | 3000 | 1.9422 | 0.2042 |
66
- | 0.0129 | 141.3 | 3250 | 1.9263 | 0.1969 |
67
- | 0.0112 | 152.17 | 3500 | 1.9063 | 0.2012 |
68
- | 0.0108 | 163.04 | 3750 | 1.8590 | 0.1978 |
69
- | 0.0095 | 173.91 | 4000 | 1.9030 | 0.2008 |
70
 
71
 
72
  ### 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: 2.3736
21
+ - Wer: 0.1858
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: 4000
48
  - mixed_precision_training: Native AMP
49
 
 
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Wer |
53
  |:-------------:|:------:|:----:|:---------------:|:------:|
54
+ | 20.81 | 10.87 | 250 | 3.8675 | 1.0 |
55
+ | 3.2992 | 21.74 | 500 | 3.1858 | 1.0 |
56
+ | 3.0813 | 32.61 | 750 | 3.0758 | 1.0 |
57
+ | 2.3718 | 43.48 | 1000 | 1.3480 | 0.8266 |
58
+ | 0.3075 | 54.35 | 1250 | 1.6498 | 0.2290 |
59
+ | 0.1081 | 65.22 | 1500 | 1.8213 | 0.2012 |
60
+ | 0.0732 | 76.09 | 1750 | 1.8933 | 0.1952 |
61
+ | 0.0517 | 86.96 | 2000 | 2.0154 | 0.2059 |
62
+ | 0.0386 | 97.83 | 2250 | 2.0444 | 0.1948 |
63
+ | 0.0323 | 108.7 | 2500 | 2.2603 | 0.2003 |
64
+ | 0.0272 | 119.57 | 2750 | 2.2578 | 0.1952 |
65
+ | 0.0234 | 130.43 | 3000 | 2.2854 | 0.1880 |
66
+ | 0.0203 | 141.3 | 3250 | 2.3553 | 0.1867 |
67
+ | 0.0181 | 152.17 | 3500 | 2.3723 | 0.1905 |
68
+ | 0.0165 | 163.04 | 3750 | 2.3793 | 0.1854 |
69
+ | 0.016 | 173.91 | 4000 | 2.3736 | 0.1858 |
70
 
71
 
72
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7cf763cefcf54f43e7c8f30efe3beb7c7d669716e17f82123e787be1568d768d
3
  size 377629528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6154ac869a41441b6230ceaf24a38587f1f5158f04472ccc1ed5c0453b94f54a
3
  size 377629528