Hugging Face's logo

MelanieKoe
/
w2v2-base-pretrained_lr5e-5_at0.6_da1

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MelanieKoe commited on
Commit
fb39889
•
1 Parent(s): 84c1da1

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: 1.4542
21
- - Wer: 0.1867
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
- | 14.9488 | 4.24 | 250 | 3.6144 | 1.0 |
55
- | 3.1468 | 8.47 | 500 | 3.2251 | 1.0 |
56
- | 2.5166 | 12.71 | 750 | 1.3839 | 0.9603 |
57
- | 0.4698 | 16.95 | 1000 | 0.6829 | 0.2909 |
58
- | 0.2122 | 21.19 | 1250 | 0.9930 | 0.2217 |
59
- | 0.1236 | 25.42 | 1500 | 1.1644 | 0.2140 |
60
- | 0.0898 | 29.66 | 1750 | 1.0494 | 0.2076 |
61
- | 0.0664 | 33.9 | 2000 | 1.1845 | 0.2093 |
62
- | 0.0521 | 38.14 | 2250 | 1.2057 | 0.2089 |
63
- | 0.0417 | 42.37 | 2500 | 1.3375 | 0.1914 |
64
- | 0.0359 | 46.61 | 2750 | 1.5455 | 0.1880 |
65
- | 0.0315 | 50.85 | 3000 | 1.3454 | 0.1884 |
66
- | 0.0267 | 55.08 | 3250 | 1.2789 | 0.1944 |
67
- | 0.0239 | 59.32 | 3500 | 1.3917 | 0.1909 |
68
- | 0.0223 | 63.56 | 3750 | 1.4291 | 0.1897 |
69
- | 0.0202 | 67.8 | 4000 | 1.4542 | 0.1867 |
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: 1.4078
21
+ - Wer: 0.1662
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
+ | 25.5608 | 4.24 | 250 | 20.1943 | 1.0 |
55
+ | 5.6218 | 8.47 | 500 | 3.4933 | 1.0 |
56
+ | 3.134 | 12.71 | 750 | 3.1332 | 1.0 |
57
+ | 3.0842 | 16.95 | 1000 | 3.1196 | 1.0 |
58
+ | 2.6714 | 21.19 | 1250 | 1.4095 | 0.9932 |
59
+ | 0.5642 | 25.42 | 1500 | 0.6028 | 0.4327 |
60
+ | 0.2724 | 29.66 | 1750 | 0.7355 | 0.2396 |
61
+ | 0.1676 | 33.9 | 2000 | 0.9693 | 0.2264 |
62
+ | 0.1247 | 38.14 | 2250 | 1.0615 | 0.2008 |
63
+ | 0.0945 | 42.37 | 2500 | 1.2612 | 0.1841 |
64
+ | 0.0784 | 46.61 | 2750 | 1.1492 | 0.1850 |
65
+ | 0.0672 | 50.85 | 3000 | 1.3113 | 0.1751 |
66
+ | 0.0567 | 55.08 | 3250 | 1.3594 | 0.1721 |
67
+ | 0.0507 | 59.32 | 3500 | 1.4195 | 0.1721 |
68
+ | 0.0478 | 63.56 | 3750 | 1.3916 | 0.1670 |
69
+ | 0.0445 | 67.8 | 4000 | 1.4078 | 0.1662 |
70
 
71
 
72
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:704537c4b2ae4e5387d23baac3b6d597e3b18ffe49c2f04a5d008fb8eb4f92c2
3
  size 377629528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:286308fb34631764696e425f6a0ca0a9138d756293e5e42249a18f875ae01ed0
3
  size 377629528