Hugging Face's logo

MelanieKoe
/
w2v2-base-pretrained_lr5e-5_at0.8_da1

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MelanieKoe commited on
Commit
4110c5a
•
1 Parent(s): 1f12945

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: 1.6691
21
- - Wer: 0.1858
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
- | 14.0341 | 5.43 | 250 | 3.5667 | 1.0 |
55
- | 3.138 | 10.87 | 500 | 3.1976 | 1.0 |
56
- | 3.0471 | 16.3 | 750 | 3.0452 | 1.0 |
57
- | 1.3116 | 21.74 | 1000 | 1.0225 | 0.3358 |
58
- | 0.1806 | 27.17 | 1250 | 1.2151 | 0.2533 |
59
- | 0.1065 | 32.61 | 1500 | 1.2673 | 0.2384 |
60
- | 0.0707 | 38.04 | 1750 | 1.4184 | 0.1888 |
61
- | 0.0547 | 43.48 | 2000 | 1.6087 | 0.1901 |
62
- | 0.0471 | 48.91 | 2250 | 1.4537 | 0.1880 |
63
- | 0.0381 | 54.35 | 2500 | 1.6858 | 0.1845 |
64
- | 0.0307 | 59.78 | 2750 | 1.5607 | 0.1961 |
65
- | 0.0269 | 65.22 | 3000 | 1.6716 | 0.1862 |
66
- | 0.024 | 70.65 | 3250 | 1.5554 | 0.1884 |
67
- | 0.0199 | 76.09 | 3500 | 1.7169 | 0.1850 |
68
- | 0.0187 | 81.52 | 3750 | 1.6661 | 0.1858 |
69
- | 0.0176 | 86.96 | 4000 | 1.6691 | 0.1858 |
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: 2.3273
21
+ - Wer: 0.1747
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
+ | 18.3928 | 5.43 | 250 | 4.5566 | 1.0 |
55
+ | 3.4003 | 10.87 | 500 | 3.2320 | 1.0 |
56
+ | 2.9641 | 16.3 | 750 | 2.5134 | 1.0 |
57
+ | 0.9965 | 21.74 | 1000 | 1.2335 | 0.4165 |
58
+ | 0.2103 | 27.17 | 1250 | 1.5883 | 0.2012 |
59
+ | 0.1252 | 32.61 | 1500 | 1.5291 | 0.1837 |
60
+ | 0.0911 | 38.04 | 1750 | 1.8433 | 0.1901 |
61
+ | 0.0717 | 43.48 | 2000 | 1.9624 | 0.1858 |
62
+ | 0.0609 | 48.91 | 2250 | 1.9417 | 0.1781 |
63
+ | 0.0488 | 54.35 | 2500 | 2.0794 | 0.1751 |
64
+ | 0.041 | 59.78 | 2750 | 2.2785 | 0.1837 |
65
+ | 0.0357 | 65.22 | 3000 | 2.1884 | 0.1709 |
66
+ | 0.0325 | 70.65 | 3250 | 2.2440 | 0.1764 |
67
+ | 0.0286 | 76.09 | 3500 | 2.2743 | 0.1764 |
68
+ | 0.0263 | 81.52 | 3750 | 2.2614 | 0.1730 |
69
+ | 0.0256 | 86.96 | 4000 | 2.3273 | 0.1747 |
70
 
71
 
72
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:76b8964eafc32c50189e78465229d5e332d59f92756bd54c7e603bf7284abd74
3
  size 377629528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8b143d6fc4df5865f92492fa284aef1d3c6a5af03f0df70331c6be1e2bf2362
3
  size 377629528