Hugging Face's logo

MelanieKoe
/
w2v2-base-pretrained_lr5e-5_at0.0_da1

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MelanieKoe commited on
Commit
242e9f6
•
1 Parent(s): 3995cf2

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: 1.0838
21
- - Wer: 0.1768
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
- | 16.3345 | 3.91 | 250 | 3.9551 | 1.0 |
55
- | 3.2558 | 7.81 | 500 | 3.1516 | 1.0 |
56
- | 2.9971 | 11.72 | 750 | 2.4403 | 1.0 |
57
- | 0.9923 | 15.62 | 1000 | 0.6040 | 0.4938 |
58
- | 0.2971 | 19.53 | 1250 | 0.6870 | 0.2828 |
59
- | 0.1765 | 23.44 | 1500 | 0.8956 | 0.2431 |
60
- | 0.1185 | 27.34 | 1750 | 0.9472 | 0.2029 |
61
- | 0.0919 | 31.25 | 2000 | 1.0306 | 0.1833 |
62
- | 0.0692 | 35.16 | 2250 | 0.9844 | 0.1939 |
63
- | 0.0577 | 39.06 | 2500 | 1.0122 | 0.1862 |
64
- | 0.0467 | 42.97 | 2750 | 1.0849 | 0.1734 |
65
- | 0.0407 | 46.88 | 3000 | 0.9989 | 0.1841 |
66
- | 0.0341 | 50.78 | 3250 | 1.0820 | 0.1875 |
67
- | 0.0299 | 54.69 | 3500 | 1.1344 | 0.1747 |
68
- | 0.0291 | 58.59 | 3750 | 1.0495 | 0.1845 |
69
- | 0.0247 | 62.5 | 4000 | 1.0838 | 0.1768 |
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: 1.2935
21
+ - Wer: 0.1709
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
+ | 22.479 | 3.91 | 250 | 4.5097 | 1.0 |
55
+ | 3.5142 | 7.81 | 500 | 3.2115 | 1.0 |
56
+ | 3.144 | 11.72 | 750 | 3.1091 | 1.0 |
57
+ | 2.5815 | 15.62 | 1000 | 1.2241 | 0.9991 |
58
+ | 0.6479 | 19.53 | 1250 | 0.5868 | 0.3328 |
59
+ | 0.3255 | 23.44 | 1500 | 0.7123 | 0.2050 |
60
+ | 0.2136 | 27.34 | 1750 | 0.8753 | 0.1854 |
61
+ | 0.1561 | 31.25 | 2000 | 0.9095 | 0.1892 |
62
+ | 0.1195 | 35.16 | 2250 | 1.0824 | 0.1828 |
63
+ | 0.0966 | 39.06 | 2500 | 1.0976 | 0.1756 |
64
+ | 0.0829 | 42.97 | 2750 | 1.1946 | 0.1734 |
65
+ | 0.0724 | 46.88 | 3000 | 1.2161 | 0.1713 |
66
+ | 0.0611 | 50.78 | 3250 | 1.2877 | 0.1739 |
67
+ | 0.0555 | 54.69 | 3500 | 1.3169 | 0.1687 |
68
+ | 0.0537 | 58.59 | 3750 | 1.2744 | 0.1764 |
69
+ | 0.0481 | 62.5 | 4000 | 1.2935 | 0.1709 |
70
 
71
 
72
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4702447f38375768dcf1d199dded778e0b60fa41f34407deaba1976ec8e9c494
3
  size 377629528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0e8959d9823a9cfa79499aa77a7808588306407c02dc5703750d536b083aff6
3
  size 377629528