Hugging Face's logo

MelanieKoe
/
w2v2-base-pretrained_lr1e-4_at0.8_da0.5

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MelanieKoe commited on
Commit
5ab9a72
•
1 Parent(s): 7782b70

Model save

Browse files
Files changed (2) hide show
  1. README.md +20 -52
  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: 2.7317
21
- - Wer: 0.1901
22
 
23
  ## Model description
24
 
@@ -43,60 +43,28 @@ 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: 100
47
- - num_epochs: 200
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
- | Training Loss | Epoch | Step | Validation Loss | Wer |
53
- |:-------------:|:------:|:----:|:---------------:|:------:|
54
- | 17.7776 | 4.35 | 100 | 3.5672 | 1.0 |
55
- | 3.1868 | 8.7 | 200 | 3.1741 | 1.0 |
56
- | 3.0965 | 13.04 | 300 | 3.1231 | 1.0 |
57
- | 3.074 | 17.39 | 400 | 3.1511 | 1.0 |
58
- | 3.0648 | 21.74 | 500 | 3.1050 | 1.0 |
59
- | 3.0507 | 26.09 | 600 | 3.0931 | 1.0 |
60
- | 2.9697 | 30.43 | 700 | 2.9625 | 1.0 |
61
- | 2.2903 | 34.78 | 800 | 1.6965 | 1.0 |
62
- | 0.6175 | 39.13 | 900 | 1.2901 | 0.5557 |
63
- | 0.2126 | 43.48 | 1000 | 1.3785 | 0.3366 |
64
- | 0.1406 | 47.83 | 1100 | 1.3940 | 0.2636 |
65
- | 0.0967 | 52.17 | 1200 | 1.8359 | 0.2469 |
66
- | 0.0827 | 56.52 | 1300 | 1.6063 | 0.2704 |
67
- | 0.0699 | 60.87 | 1400 | 2.0679 | 0.2465 |
68
- | 0.0579 | 65.22 | 1500 | 2.0735 | 0.2311 |
69
- | 0.056 | 69.57 | 1600 | 2.0033 | 0.2191 |
70
- | 0.0435 | 73.91 | 1700 | 2.0973 | 0.2247 |
71
- | 0.0389 | 78.26 | 1800 | 2.0766 | 0.2226 |
72
- | 0.0334 | 82.61 | 1900 | 2.2471 | 0.2140 |
73
- | 0.0296 | 86.96 | 2000 | 2.2653 | 0.2102 |
74
- | 0.0291 | 91.3 | 2100 | 2.2845 | 0.2448 |
75
- | 0.0246 | 95.65 | 2200 | 2.1763 | 0.2110 |
76
- | 0.0218 | 100.0 | 2300 | 2.5201 | 0.1991 |
77
- | 0.0227 | 104.35 | 2400 | 2.2250 | 0.2196 |
78
- | 0.0194 | 108.7 | 2500 | 2.4203 | 0.2179 |
79
- | 0.0191 | 113.04 | 2600 | 2.4685 | 0.2016 |
80
- | 0.0195 | 117.39 | 2700 | 2.4752 | 0.2089 |
81
- | 0.0155 | 121.74 | 2800 | 2.3154 | 0.2114 |
82
- | 0.015 | 126.09 | 2900 | 2.2900 | 0.2076 |
83
- | 0.0143 | 130.43 | 3000 | 2.5405 | 0.2016 |
84
- | 0.0129 | 134.78 | 3100 | 2.6555 | 0.2025 |
85
- | 0.0139 | 139.13 | 3200 | 2.4805 | 0.2050 |
86
- | 0.0102 | 143.48 | 3300 | 2.5350 | 0.1956 |
87
- | 0.0119 | 147.83 | 3400 | 2.4560 | 0.2127 |
88
- | 0.0099 | 152.17 | 3500 | 2.5115 | 0.1974 |
89
- | 0.0096 | 156.52 | 3600 | 2.7533 | 0.2046 |
90
- | 0.0086 | 160.87 | 3700 | 2.6834 | 0.1982 |
91
- | 0.0085 | 165.22 | 3800 | 2.6764 | 0.1986 |
92
- | 0.0089 | 169.57 | 3900 | 2.6680 | 0.1914 |
93
- | 0.0073 | 173.91 | 4000 | 2.6164 | 0.1948 |
94
- | 0.0081 | 178.26 | 4100 | 2.6260 | 0.1922 |
95
- | 0.0071 | 182.61 | 4200 | 2.6611 | 0.1901 |
96
- | 0.0068 | 186.96 | 4300 | 2.7081 | 0.1914 |
97
- | 0.007 | 191.3 | 4400 | 2.7668 | 0.1892 |
98
- | 0.0061 | 195.65 | 4500 | 2.7364 | 0.1888 |
99
- | 0.0063 | 200.0 | 4600 | 2.7317 | 0.1901 |
100
 
101
 
102
  ### 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: 12.5971
21
+ - Wer: 1.0
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: 3500
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
53
+ |:-------------:|:------:|:----:|:---------------:|:---:|
54
+ | 22.3588 | 10.87 | 250 | 23.6457 | 1.0 |
55
+ | 8.9908 | 21.74 | 500 | 12.4169 | 1.0 |
56
+ | 11.1075 | 32.61 | 750 | 20.5658 | 1.0 |
57
+ | 13.3015 | 43.48 | 1000 | 20.3357 | 1.0 |
58
+ | 10.7105 | 54.35 | 1250 | 12.4533 | 1.0 |
59
+ | 8.8951 | 65.22 | 1500 | 12.5400 | 1.0 |
60
+ | 8.8229 | 76.09 | 1750 | 12.5416 | 1.0 |
61
+ | 8.7878 | 86.96 | 2000 | 12.5533 | 1.0 |
62
+ | 8.8351 | 97.83 | 2250 | 12.5107 | 1.0 |
63
+ | 8.8182 | 108.7 | 2500 | 12.6065 | 1.0 |
64
+ | 8.7863 | 119.57 | 2750 | 12.6950 | 1.0 |
65
+ | 8.801 | 130.43 | 3000 | 12.6095 | 1.0 |
66
+ | 8.8417 | 141.3 | 3250 | 12.6030 | 1.0 |
67
+ | 8.7401 | 152.17 | 3500 | 12.5971 | 1.0 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
68
 
69
 
70
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3ceb6f9b4fdd550edd532974d23b917480fcad85a0da2620095355a808e88ecf
3
  size 377629528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26fcae8834a65ee8b15a20a7a8e82f9880e3352c1f7f9ca2dea2f15f52d34cbb
3
  size 377629528