Hugging Face's logo

MelanieKoe
/
w2v2-base-pretrained_lr5e-5_at0.8_da0.2

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MelanieKoe commited on
Commit
2a224e0
·
verified ·
1 Parent(s): 50f05ff

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: 1.6497
21
- - Wer: 0.2191
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
- | 19.3673 | 25.0 | 250 | 3.5485 | 1.0 |
55
- | 3.1951 | 50.0 | 500 | 3.0982 | 1.0 |
56
- | 2.9571 | 75.0 | 750 | 2.5890 | 1.0 |
57
- | 0.8132 | 100.0 | 1000 | 1.0110 | 0.4161 |
58
- | 0.0936 | 125.0 | 1250 | 1.4372 | 0.2636 |
59
- | 0.0487 | 150.0 | 1500 | 1.4519 | 0.2533 |
60
- | 0.0314 | 175.0 | 1750 | 1.5135 | 0.2554 |
61
- | 0.0204 | 200.0 | 2000 | 1.6471 | 0.2439 |
62
- | 0.0189 | 225.0 | 2250 | 1.5987 | 0.2405 |
63
- | 0.013 | 250.0 | 2500 | 1.7187 | 0.2251 |
64
- | 0.0109 | 275.0 | 2750 | 1.6743 | 0.2302 |
65
- | 0.0099 | 300.0 | 3000 | 1.6391 | 0.2200 |
66
- | 0.0084 | 325.0 | 3250 | 1.6393 | 0.2200 |
67
- | 0.0074 | 350.0 | 3500 | 1.7161 | 0.2136 |
68
- | 0.0066 | 375.0 | 3750 | 1.6937 | 0.2170 |
69
- | 0.0063 | 400.0 | 4000 | 1.6497 | 0.2191 |
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: 2.3279
21
+ - Wer: 0.2674
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
+ | 25.9775 | 25.0 | 250 | 4.0805 | 1.0 |
55
+ | 3.4724 | 50.0 | 500 | 3.1777 | 1.0 |
56
+ | 3.0775 | 75.0 | 750 | 3.0307 | 1.0 |
57
+ | 2.7225 | 100.0 | 1000 | 2.2425 | 1.0 |
58
+ | 0.8725 | 125.0 | 1250 | 1.5857 | 0.4584 |
59
+ | 0.1745 | 150.0 | 1500 | 1.8015 | 0.3503 |
60
+ | 0.0937 | 175.0 | 1750 | 1.8837 | 0.3131 |
61
+ | 0.0645 | 200.0 | 2000 | 2.0461 | 0.2947 |
62
+ | 0.0528 | 225.0 | 2250 | 2.1465 | 0.2853 |
63
+ | 0.0366 | 250.0 | 2500 | 2.1419 | 0.2875 |
64
+ | 0.0336 | 275.0 | 2750 | 2.1437 | 0.2755 |
65
+ | 0.026 | 300.0 | 3000 | 2.1733 | 0.2717 |
66
+ | 0.0247 | 325.0 | 3250 | 2.2546 | 0.2683 |
67
+ | 0.0207 | 350.0 | 3500 | 2.2603 | 0.2700 |
68
+ | 0.0185 | 375.0 | 3750 | 2.3492 | 0.2670 |
69
+ | 0.0186 | 400.0 | 4000 | 2.3279 | 0.2674 |
70
 
71
 
72
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8be6589a156ec8fdd0915ec5cd04818b1ab135c5939b558121ea38eb61f2286a
3
  size 377629528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3f6ee9cfaf6f431e6cb2f140d8bd995c03f9c63ca52c80b3a5e10aa2bfb5df3
3
  size 377629528