Hugging Face's logo

mozilla-ai
/
whisper-small-sv

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Swedish
whisper
Eval Results (legacy)
Model card Files
xet
 Metrics Community
kostissz commited on
Commit
4680695
·
verified ·
1 Parent(s): 8e65386

End of training

Browse files
Files changed (2) hide show
  1. README.md +20 -33
  2. runs/Mar06_11-07-10_gpu-pod/events.out.tfevents.1741266944.gpu-pod.61799.1 +3 -0
README.md CHANGED
@@ -23,7 +23,7 @@ model-index:
23
  metrics:
24
  - name: Wer
25
  type: wer
26
- value: 20.14691254112786
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -33,9 +33,9 @@ should probably proofread and complete it, then remove this comment. -->
33
 
34
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the common_voice_17_0 dataset.
35
  It achieves the following results on the evaluation set:
36
- - Loss: 0.3111
37
  - Model Preparation Time: 0.0044
38
- - Wer: 20.1469
39
 
40
  ## Model description
41
 
@@ -55,44 +55,31 @@ More information needed
55
 
56
  The following hyperparameters were used during training:
57
  - learning_rate: 1e-05
58
- - train_batch_size: 258
59
- - eval_batch_size: 64
60
  - seed: 42
61
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
62
  - lr_scheduler_type: linear
63
  - lr_scheduler_warmup_steps: 50
64
- - training_steps: 1250
65
  - mixed_precision_training: Native AMP
66
 
67
  ### Training results
68
 
69
- | Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Wer |
70
- |:-------------:|:-------:|:----:|:---------------:|:----------------------:|:-------:|
71
- | 0.6188 | 0.9804 | 50 | 0.3424 | 0.0044 | 23.1107 |
72
- | 0.2417 | 1.9608 | 100 | 0.3089 | 0.0044 | 21.0881 |
73
- | 0.1456 | 2.9412 | 150 | 0.3038 | 0.0044 | 20.7412 |
74
- | 0.0928 | 3.9216 | 200 | 0.3111 | 0.0044 | 20.1469 |
75
- | 0.0554 | 4.9020 | 250 | 0.3231 | 0.0044 | 20.3739 |
76
- | 0.035 | 5.8824 | 300 | 0.3367 | 0.0044 | 20.8432 |
77
- | 0.0215 | 6.8627 | 350 | 0.3619 | 0.0044 | 20.8458 |
78
- | 0.0143 | 7.8431 | 400 | 0.3768 | 0.0044 | 20.9299 |
79
- | 0.0101 | 8.8235 | 450 | 0.3880 | 0.0044 | 20.8509 |
80
- | 0.0084 | 9.8039 | 500 | 0.3960 | 0.0044 | 20.8993 |
81
- | 0.0072 | 10.7843 | 550 | 0.3999 | 0.0044 | 21.0014 |
82
- | 0.0059 | 11.7647 | 600 | 0.4069 | 0.0044 | 20.8942 |
83
- | 0.0053 | 12.7451 | 650 | 0.4130 | 0.0044 | 20.9656 |
84
- | 0.0047 | 13.7255 | 700 | 0.4177 | 0.0044 | 20.9963 |
85
- | 0.0043 | 14.7059 | 750 | 0.4208 | 0.0044 | 20.9478 |
86
- | 0.004 | 15.6863 | 800 | 0.4241 | 0.0044 | 21.0371 |
87
- | 0.0037 | 16.6667 | 850 | 0.4265 | 0.0044 | 21.0600 |
88
- | 0.0035 | 17.6471 | 900 | 0.4298 | 0.0044 | 21.1034 |
89
- | 0.0034 | 18.6275 | 950 | 0.4317 | 0.0044 | 21.0983 |
90
- | 0.0032 | 19.6078 | 1000 | 0.4334 | 0.0044 | 21.1416 |
91
- | 0.0031 | 20.5882 | 1050 | 0.4351 | 0.0044 | 21.1518 |
92
- | 0.003 | 21.5686 | 1100 | 0.4361 | 0.0044 | 21.1748 |
93
- | 0.0029 | 22.5490 | 1150 | 0.4368 | 0.0044 | 21.1620 |
94
- | 0.0029 | 23.5294 | 1200 | 0.4374 | 0.0044 | 21.1799 |
95
- | 0.0029 | 24.5098 | 1250 | 0.4377 | 0.0044 | 21.1722 |
96
 
97
 
98
  ### Framework versions
 
23
  metrics:
24
  - name: Wer
25
  type: wer
26
+ value: 20.88912694161757
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
33
 
34
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the common_voice_17_0 dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.3054
37
  - Model Preparation Time: 0.0044
38
+ - Wer: 20.8891
39
 
40
  ## Model description
41
 
 
55
 
56
  The following hyperparameters were used during training:
57
  - learning_rate: 1e-05
58
+ - train_batch_size: 64
59
+ - eval_batch_size: 32
60
  - seed: 42
61
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
62
  - lr_scheduler_type: linear
63
  - lr_scheduler_warmup_steps: 50
64
+ - training_steps: 300
65
  - mixed_precision_training: Native AMP
66
 
67
  ### Training results
68
 
69
+ | Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Wer |
70
+ |:-------------:|:------:|:----:|:---------------:|:----------------------:|:-------:|
71
+ | 0.9087 | 0.1232 | 25 | 0.5954 | 0.0044 | 25.6485 |
72
+ | 0.3769 | 0.2463 | 50 | 0.3614 | 0.0044 | 23.9728 |
73
+ | 0.3282 | 0.3695 | 75 | 0.3457 | 0.0044 | 23.8478 |
74
+ | 0.3236 | 0.4926 | 100 | 0.3340 | 0.0044 | 22.8939 |
75
+ | 0.3075 | 0.6158 | 125 | 0.3260 | 0.0044 | 22.5853 |
76
+ | 0.2922 | 0.7389 | 150 | 0.3186 | 0.0044 | 21.8711 |
77
+ | 0.287 | 0.8621 | 175 | 0.3140 | 0.0044 | 21.6670 |
78
+ | 0.2845 | 0.9852 | 200 | 0.3093 | 0.0044 | 21.4196 |
79
+ | 0.195 | 1.1084 | 225 | 0.3080 | 0.0044 | 21.3610 |
80
+ | 0.1679 | 1.2315 | 250 | 0.3079 | 0.0044 | 21.1059 |
81
+ | 0.1726 | 1.3547 | 275 | 0.3060 | 0.0044 | 21.0243 |
82
+ | 0.165 | 1.4778 | 300 | 0.3054 | 0.0044 | 20.8891 |
 
 
 
 
 
 
 
 
 
 
 
 
 
83
 
84
 
85
  ### Framework versions
runs/Mar06_11-07-10_gpu-pod/events.out.tfevents.1741266944.gpu-pod.61799.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c68925fcd35a0c1e19c08ecdbb0a4695280768b061c28d3325c028af49b3477
3
+ size 472