rmacek commited on
Commit
7c120f0
·
verified ·
1 Parent(s): fe881eb

End of training

Browse files
Files changed (2) hide show
  1. README.md +10 -10
  2. adapter_model.safetensors +1 -1
README.md CHANGED
@@ -17,12 +17,12 @@ model-index:
17
  type: automatic-speech-recognition
18
  name: Automatic Speech Recognition
19
  dataset:
20
- name: ZIB2 Common Voice
21
  type: rmacek/ORF-whisper-small
22
  args: 'config: de, split: test'
23
  metrics:
24
  - type: wer
25
- value: 28.39606613864229
26
  name: Wer
27
  ---
28
 
@@ -31,10 +31,10 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  # Whisper ORF Bundeslaender
33
 
34
- This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the ZIB2 Common Voice dataset.
35
  It achieves the following results on the evaluation set:
36
- - Loss: 0.6318
37
- - Wer: 28.3961
38
 
39
  ## Model description
40
 
@@ -53,7 +53,7 @@ More information needed
53
  ### Training hyperparameters
54
 
55
  The following hyperparameters were used during training:
56
- - learning_rate: 1e-05
57
  - train_batch_size: 16
58
  - eval_batch_size: 8
59
  - seed: 42
@@ -67,10 +67,10 @@ The following hyperparameters were used during training:
67
 
68
  | Training Loss | Epoch | Step | Validation Loss | Wer |
69
  |:-------------:|:------:|:----:|:---------------:|:-------:|
70
- | 0.6514 | 1.7153 | 1000 | 0.6613 | 37.4578 |
71
- | 0.5883 | 3.4305 | 2000 | 0.6400 | 37.4889 |
72
- | 0.6529 | 5.1458 | 3000 | 0.6335 | 35.9455 |
73
- | 0.5928 | 6.8611 | 4000 | 0.6318 | 28.3961 |
74
 
75
 
76
  ### Framework versions
 
17
  type: automatic-speech-recognition
18
  name: Automatic Speech Recognition
19
  dataset:
20
+ name: ORF Bundesländer Voices
21
  type: rmacek/ORF-whisper-small
22
  args: 'config: de, split: test'
23
  metrics:
24
  - type: wer
25
+ value: 26.328635352109302
26
  name: Wer
27
  ---
28
 
 
31
 
32
  # Whisper ORF Bundeslaender
33
 
34
+ This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the ORF Bundesländer Voices dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.6154
37
+ - Wer: 26.3286
38
 
39
  ## Model description
40
 
 
53
  ### Training hyperparameters
54
 
55
  The following hyperparameters were used during training:
56
+ - learning_rate: 0.0001
57
  - train_batch_size: 16
58
  - eval_batch_size: 8
59
  - seed: 42
 
67
 
68
  | Training Loss | Epoch | Step | Validation Loss | Wer |
69
  |:-------------:|:------:|:----:|:---------------:|:-------:|
70
+ | 0.5731 | 1.7153 | 1000 | 0.6095 | 36.1178 |
71
+ | 0.4196 | 3.4305 | 2000 | 0.6040 | 26.1875 |
72
+ | 0.4214 | 5.1458 | 3000 | 0.6086 | 26.3813 |
73
+ | 0.3579 | 6.8611 | 4000 | 0.6154 | 26.3286 |
74
 
75
 
76
  ### Framework versions
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8ae1c1b6961fb627b73eb2cbf741bcd568605da3e70936be5f91882bc77bc29d
3
  size 14176064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ddfacbca5783de5d892f95f08e49c0fa4e13a26e32f62be8d86f212827b43ba
3
  size 14176064