greenw0lf commited on
Commit
eada0e6
1 Parent(s): 92a452a

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +26 -3
README.md CHANGED
@@ -22,6 +22,19 @@ model-index:
22
  - name: Wer
23
  type: wer
24
  value: 0.040494215112126836
 
 
 
 
 
 
 
 
 
 
 
 
 
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -34,20 +47,30 @@ It achieves the following results on the evaluation set:
34
  - Loss: 0.0465
35
  - Wer: 0.0405
36
 
 
 
 
 
37
  ## Model description
38
 
39
- More information needed
 
 
40
 
41
  ## Intended uses & limitations
42
 
43
- More information needed
 
 
44
 
45
  ## Training and evaluation data
46
 
47
- More information needed
48
 
49
  ## Training procedure
50
 
 
 
51
  ### Training hyperparameters
52
 
53
  The following hyperparameters were used during training:
 
22
  - name: Wer
23
  type: wer
24
  value: 0.040494215112126836
25
+ - task:
26
+ name: Automatic Speech Recognition
27
+ type: automatic-speech-recognition
28
+ dataset:
29
+ name: common_voice_8_0
30
+ type: common_voice_8_0
31
+ config: fy-NL
32
+ split: test
33
+ args: fy-NL
34
+ metrics:
35
+ - name: Wer
36
+ type: wer
37
+ value: 0.04223876282699812
38
  ---
39
 
40
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
47
  - Loss: 0.0465
48
  - Wer: 0.0405
49
 
50
+ And on the test set:
51
+
52
+ - Wer: 0.0422
53
+
54
  ## Model description
55
 
56
+ This model has been developed for my Master's thesis in "Voice Technology" at Rijksuniversiteit Groningen - Campus Fryslân. It corresponds to experiment 7 where
57
+ I use as training set all validated data (~ 50 hours) except the test and evaluation sets (~ 4.5 hours each).
58
+ The number of training hours adds up to 41 hours of Frisian speech. This varies from experiment 2 because I fine-tune on the 2B parameters version of XLS-R.
59
 
60
  ## Intended uses & limitations
61
 
62
+ The intended use is for recognizing Frisian speech.
63
+
64
+ Limitations include no LM rescoring and using version 8.0 of Common Voice instead of 13.0.
65
 
66
  ## Training and evaluation data
67
 
68
+ The evaluation split used is the one available in the Common Voice 8.0 Frisian subset. The train split corresponds to all of the validated data except for the recordings found in the evaluation and test splits.
69
 
70
  ## Training procedure
71
 
72
+ The script used for training this model can be found in this GitHub repository: [link](https://github.com/greenw0lf/MSc-VT-Thesis/).
73
+
74
  ### Training hyperparameters
75
 
76
  The following hyperparameters were used during training: