Hugging Face's logo

marinone94
/
xls-r-300m-sv-robust

Automatic Speech Recognition
Transformers
PyTorch
Swedish
wav2vec2
mozilla-foundation/common_voice_9_0
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
1
marinone94 commited on
Commit
55a2553
·
1 Parent(s): 05cee88

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +54 -48
README.md CHANGED
@@ -1,11 +1,9 @@
1
  ---
2
  license: cc0-1.0
3
  tags:
4
- - automatic-speech-recognition
5
- - marinone94/nst_sv
6
  - generated_from_trainer
7
  datasets:
8
- - nst_sv
9
  model-index:
10
  - name: ''
11
  results: []
@@ -16,10 +14,10 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  #
18
 
19
- This model is a fine-tuned version of [KBLab/wav2vec2-large-voxrex](https://huggingface.co/KBLab/wav2vec2-large-voxrex) on the MARINONE94/NST_SV - DISTANT_CHANNEL dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: inf
22
- - Wer: 1.0
23
 
24
  ## Model description
25
 
@@ -38,7 +36,7 @@ More information needed
38
  ### Training hyperparameters
39
 
40
  The following hyperparameters were used during training:
41
- - learning_rate: 0.00075
42
  - train_batch_size: 32
43
  - eval_batch_size: 32
44
  - seed: 42
@@ -46,51 +44,59 @@ The following hyperparameters were used during training:
46
  - total_train_batch_size: 128
47
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
  - lr_scheduler_type: linear
49
- - lr_scheduler_warmup_ratio: 0.02
50
- - num_epochs: 2.0
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
- | Training Loss | Epoch | Step | Validation Loss | Wer |
56
- |:-------------:|:-----:|:----:|:---------------:|:---:|
57
- | 3.4039 | 0.05 | 100 | inf | 1.0 |
58
- | 3.4396 | 0.11 | 200 | inf | 1.0 |
59
- | 3.483 | 0.16 | 300 | inf | 1.0 |
60
- | 3.5014 | 0.21 | 400 | inf | 1.0 |
61
- | 3.331 | 0.27 | 500 | inf | 1.0 |
62
- | 3.4809 | 0.32 | 600 | inf | 1.0 |
63
- | 3.4678 | 0.37 | 700 | inf | 1.0 |
64
- | 3.4596 | 0.43 | 800 | inf | 1.0 |
65
- | 3.4644 | 0.48 | 900 | inf | 1.0 |
66
- | 3.4671 | 0.53 | 1000 | inf | 1.0 |
67
- | 3.6005 | 0.59 | 1100 | inf | 1.0 |
68
- | 3.9182 | 0.64 | 1200 | inf | 1.0 |
69
- | 3.6466 | 0.69 | 1300 | inf | 1.0 |
70
- | 3.6932 | 0.75 | 1400 | inf | 1.0 |
71
- | 3.7939 | 0.8 | 1500 | inf | 1.0 |
72
- | 3.9284 | 0.85 | 1600 | inf | 1.0 |
73
- | 3.7859 | 0.91 | 1700 | inf | 1.0 |
74
- | 3.9363 | 0.96 | 1800 | inf | 1.0 |
75
- | 3.7573 | 1.01 | 1900 | inf | 1.0 |
76
- | 3.7553 | 1.07 | 2000 | inf | 1.0 |
77
- | 3.7606 | 1.12 | 2100 | inf | 1.0 |
78
- | 3.7514 | 1.17 | 2200 | inf | 1.0 |
79
- | 3.7472 | 1.23 | 2300 | inf | 1.0 |
80
- | 3.7478 | 1.28 | 2400 | inf | 1.0 |
81
- | 3.7496 | 1.33 | 2500 | inf | 1.0 |
82
- | 3.7513 | 1.39 | 2600 | inf | 1.0 |
83
- | 3.7497 | 1.44 | 2700 | inf | 1.0 |
84
- | 3.7539 | 1.49 | 2800 | inf | 1.0 |
85
- | 3.7581 | 1.55 | 2900 | inf | 1.0 |
86
- | 3.7572 | 1.6 | 3000 | inf | 1.0 |
87
- | 3.7589 | 1.66 | 3100 | inf | 1.0 |
88
- | 3.7592 | 1.71 | 3200 | inf | 1.0 |
89
- | 3.7531 | 1.76 | 3300 | inf | 1.0 |
90
- | 3.7567 | 1.82 | 3400 | inf | 1.0 |
91
- | 3.7613 | 1.87 | 3500 | inf | 1.0 |
92
- | 3.7516 | 1.92 | 3600 | inf | 1.0 |
93
- | 3.7581 | 1.98 | 3700 | inf | 1.0 |
 
 
 
 
 
 
 
 
94
 
95
 
96
  ### Framework versions
 
1
  ---
2
  license: cc0-1.0
3
  tags:
 
 
4
  - generated_from_trainer
5
  datasets:
6
+ - common_voice
7
  model-index:
8
  - name: ''
9
  results: []
 
14
 
15
  #
16
 
17
+ This model is a fine-tuned version of [marinone94/xls-r-300m-sv-robust](https://huggingface.co/marinone94/xls-r-300m-sv-robust) on the common_voice dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.1501
20
+ - Wer: 0.1265
21
 
22
  ## Model description
23
 
 
36
  ### Training hyperparameters
37
 
38
  The following hyperparameters were used during training:
39
+ - learning_rate: 0.00025
40
  - train_batch_size: 32
41
  - eval_batch_size: 32
42
  - seed: 42
 
44
  - total_train_batch_size: 128
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 2000
48
+ - num_epochs: 50.0
49
  - mixed_precision_training: Native AMP
50
 
51
  ### Training results
52
 
53
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
54
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
55
+ | 3.3533 | 1.1 | 100 | 3.2807 | 1.0 |
56
+ | 3.1709 | 2.2 | 200 | 3.1325 | 1.0 |
57
+ | 3.0573 | 3.3 | 300 | 3.0615 | 1.0 |
58
+ | 3.0314 | 4.39 | 400 | 3.0990 | 1.0 |
59
+ | 3.0129 | 5.49 | 500 | 3.0400 | 1.0 |
60
+ | 2.9964 | 6.59 | 600 | 2.9990 | 1.0 |
61
+ | 2.9602 | 7.69 | 700 | 2.9620 | 1.0 |
62
+ | 2.8756 | 8.79 | 800 | 2.7302 | 1.0 |
63
+ | 2.2931 | 9.89 | 900 | 1.5058 | 0.9776 |
64
+ | 1.8427 | 10.98 | 1000 | 0.9155 | 0.7832 |
65
+ | 1.4286 | 12.09 | 1100 | 0.4075 | 0.3796 |
66
+ | 1.2229 | 13.19 | 1200 | 0.2893 | 0.2652 |
67
+ | 1.1106 | 14.28 | 1300 | 0.2469 | 0.2254 |
68
+ | 1.0663 | 15.38 | 1400 | 0.2219 | 0.1973 |
69
+ | 1.0667 | 16.48 | 1500 | 0.2129 | 0.1894 |
70
+ | 1.0193 | 17.58 | 1600 | 0.1991 | 0.1789 |
71
+ | 0.9816 | 18.68 | 1700 | 0.1940 | 0.1801 |
72
+ | 0.9814 | 19.78 | 1800 | 0.1860 | 0.1667 |
73
+ | 0.9787 | 20.87 | 1900 | 0.1888 | 0.1642 |
74
+ | 0.9699 | 21.97 | 2000 | 0.1875 | 0.1704 |
75
+ | 0.9616 | 23.08 | 2100 | 0.1802 | 0.1617 |
76
+ | 0.9378 | 24.17 | 2200 | 0.1793 | 0.1577 |
77
+ | 0.888 | 25.27 | 2300 | 0.1764 | 0.1545 |
78
+ | 0.8942 | 26.37 | 2400 | 0.1674 | 0.1492 |
79
+ | 0.8701 | 27.47 | 2500 | 0.1739 | 0.1512 |
80
+ | 0.8555 | 28.57 | 2600 | 0.1690 | 0.1446 |
81
+ | 0.8513 | 29.67 | 2700 | 0.1649 | 0.1477 |
82
+ | 0.8659 | 30.77 | 2800 | 0.1637 | 0.1422 |
83
+ | 0.8419 | 31.86 | 2900 | 0.1614 | 0.1397 |
84
+ | 0.8491 | 32.96 | 3000 | 0.1595 | 0.1401 |
85
+ | 0.8395 | 34.07 | 3100 | 0.1607 | 0.1376 |
86
+ | 0.83 | 35.16 | 3200 | 0.1538 | 0.1379 |
87
+ | 0.7835 | 36.26 | 3300 | 0.1602 | 0.1408 |
88
+ | 0.7703 | 37.36 | 3400 | 0.1601 | 0.1369 |
89
+ | 0.7474 | 38.46 | 3500 | 0.1514 | 0.1342 |
90
+ | 0.7719 | 39.56 | 3600 | 0.1593 | 0.1353 |
91
+ | 0.7638 | 40.66 | 3700 | 0.1536 | 0.1338 |
92
+ | 0.771 | 41.75 | 3800 | 0.1531 | 0.1317 |
93
+ | 0.7594 | 42.85 | 3900 | 0.1498 | 0.1288 |
94
+ | 0.7383 | 43.95 | 4000 | 0.1527 | 0.1300 |
95
+ | 0.7565 | 45.05 | 4100 | 0.1482 | 0.1289 |
96
+ | 0.7697 | 46.15 | 4200 | 0.1495 | 0.1272 |
97
+ | 0.7194 | 47.25 | 4300 | 0.1493 | 0.1269 |
98
+ | 0.7479 | 48.35 | 4400 | 0.1490 | 0.1276 |
99
+ | 0.7132 | 49.45 | 4500 | 0.1501 | 0.1265 |
100
 
101
 
102
  ### Framework versions