Hugging Face's logo

deepdml
/
whisper-tiny-af-mix-norm

TensorBoard
Safetensors
Afrikaans
whisper
Generated from Trainer
Eval Results (legacy)
Model card Files
xet
Metrics Community
deepdml commited on
Commit
9ac9bfd
·
1 Parent(s): 2a9bc76

End of training

Browse files
Files changed (1) hide show
  1. README.md +70 -42
README.md CHANGED
@@ -6,10 +6,10 @@ base_model: openai/whisper-tiny
6
  tags:
7
  - generated_from_trainer
8
  datasets:
 
9
  - andreoosthuizen/afrikaans-30s
10
  - voice-biomarkers/openslr-32-hq-SA-languages-Afrikaans
11
  - google/fleurs
12
- - dsfsi-anv/multilingual-nchlt-dataset
13
  metrics:
14
  - wer
15
  model-index:
@@ -20,15 +20,16 @@ model-index:
20
  type: automatic-speech-recognition
21
  dataset:
22
  name: Common Voice 17.0
23
- type: andreoosthuizen/afrikaans-30s
24
  config: af_za
25
  split: test
26
  args: af_za
27
  metrics:
28
  - name: Wer
29
  type: wer
30
- value: 44.935064935064936
31
  ---
 
32
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
33
  should probably proofread and complete it, then remove this comment. -->
34
 
@@ -36,9 +37,9 @@ should probably proofread and complete it, then remove this comment. -->
36
 
37
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice 17.0 dataset.
38
  It achieves the following results on the evaluation set:
39
- - Loss: 1.1668
40
- - Wer: 44.9351
41
- - Cer: 18.2741
42
 
43
  ## Model description
44
 
@@ -64,32 +65,72 @@ The following hyperparameters were used during training:
64
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
65
  - lr_scheduler_type: linear
66
  - lr_scheduler_warmup_ratio: 0.04
67
- - training_steps: 2000
68
 
69
  ### Training results
70
 
71
- | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
72
- |:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|
73
- | 1.3743 | 0.05 | 100 | 1.5130 | 63.5498 | 26.0798 |
74
- | 0.7811 | 0.1 | 200 | 1.2491 | 51.7403 | 19.9455 |
75
- | 0.5477 | 0.15 | 300 | 1.1820 | 48.1732 | 18.7168 |
76
- | 0.4289 | 0.2 | 400 | 1.1518 | 49.6277 | 19.5203 |
77
- | 0.3573 | 0.25 | 500 | 1.1410 | 48.6234 | 19.9044 |
78
- | 0.2835 | 0.3 | 600 | 1.1289 | 47.3074 | 19.3649 |
79
- | 0.2602 | 0.35 | 700 | 1.1318 | 45.7835 | 19.3150 |
80
- | 0.217 | 0.4 | 800 | 1.1297 | 46.8398 | 19.1361 |
81
- | 0.2007 | 0.45 | 900 | 1.1358 | 47.4286 | 20.3296 |
82
- | 0.1798 | 0.5 | 1000 | 1.1383 | 47.4459 | 20.5906 |
83
- | 0.1548 | 0.55 | 1100 | 1.1497 | 49.3853 | 21.7723 |
84
- | 0.1384 | 0.6 | 1200 | 1.1525 | 48.5022 | 20.2827 |
85
- | 0.1325 | 0.65 | 1300 | 1.1574 | 48.8831 | 20.3120 |
86
- | 0.1259 | 0.7 | 1400 | 1.1625 | 45.4372 | 18.7637 |
87
- | 0.125 | 0.75 | 1500 | 1.1606 | 44.7100 | 18.1128 |
88
- | 0.1083 | 0.8 | 1600 | 1.1609 | 48.0519 | 20.1507 |
89
- | 0.1169 | 0.85 | 1700 | 1.1660 | 47.6017 | 19.6112 |
90
- | 0.1008 | 0.9 | 1800 | 1.1644 | 47.8095 | 19.9924 |
91
- | 0.1016 | 0.95 | 1900 | 1.1658 | 44.7273 | 18.1626 |
92
- | 0.0983 | 1.0 | 2000 | 1.1668 | 44.9351 | 18.2741 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
93
 
94
 
95
  ### Framework versions
@@ -98,16 +139,3 @@ The following hyperparameters were used during training:
98
  - Pytorch 2.3.0+cu121
99
  - Datasets 2.19.1
100
  - Tokenizers 0.19.1
101
-
102
- ## Citation
103
-
104
- Please cite the model using the following BibTeX entry:
105
-
106
- ```bibtex
107
- @misc{deepdml/whisper-tiny-af-mix-norm,
108
- title={Fine-tuned Whisper tiny ASR model for speech recognition in Afrikaans},
109
- author={Jimenez, David},
110
- howpublished={\url{https://huggingface.co/deepdml/whisper-tiny-af-mix-norm}},
111
- year={2026}
112
- }
113
- ```
 
6
  tags:
7
  - generated_from_trainer
8
  datasets:
9
+ - dsfsi-anv/multilingual-nchlt-dataset
10
  - andreoosthuizen/afrikaans-30s
11
  - voice-biomarkers/openslr-32-hq-SA-languages-Afrikaans
12
  - google/fleurs
 
13
  metrics:
14
  - wer
15
  model-index:
 
20
  type: automatic-speech-recognition
21
  dataset:
22
  name: Common Voice 17.0
23
+ type: dsfsi-anv/multilingual-nchlt-dataset
24
  config: af_za
25
  split: test
26
  args: af_za
27
  metrics:
28
  - name: Wer
29
  type: wer
30
+ value: 45.41991341991342
31
  ---
32
+
33
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
34
  should probably proofread and complete it, then remove this comment. -->
35
 
 
37
 
38
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice 17.0 dataset.
39
  It achieves the following results on the evaluation set:
40
+ - Loss: 1.2882
41
+ - Wer: 45.4199
42
+ - Cer: 18.4705
43
 
44
  ## Model description
45
 
 
65
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
66
  - lr_scheduler_type: linear
67
  - lr_scheduler_warmup_ratio: 0.04
68
+ - training_steps: 6000
69
 
70
  ### Training results
71
 
72
+ | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
73
+ |:-------------:|:------:|:----:|:---------------:|:-------:|:-------:|
74
+ | 1.9215 | 0.0167 | 100 | 1.9257 | 78.9091 | 32.3285 |
75
+ | 1.0865 | 0.0333 | 200 | 1.3997 | 56.0173 | 23.8601 |
76
+ | 0.6944 | 0.05 | 300 | 1.2310 | 50.6494 | 19.9073 |
77
+ | 0.5029 | 0.0667 | 400 | 1.1707 | 47.9827 | 18.4881 |
78
+ | 0.3939 | 0.0833 | 500 | 1.1457 | 46.8052 | 18.5086 |
79
+ | 0.3002 | 0.1 | 600 | 1.1337 | 46.0260 | 18.7901 |
80
+ | 0.2643 | 0.1167 | 700 | 1.1339 | 47.4286 | 20.2534 |
81
+ | 0.2131 | 0.1333 | 800 | 1.1313 | 48.2251 | 20.2915 |
82
+ | 0.1877 | 0.15 | 900 | 1.1379 | 50.0606 | 22.6930 |
83
+ | 0.1619 | 0.1667 | 1000 | 1.1368 | 45.9048 | 19.5613 |
84
+ | 0.1339 | 0.1833 | 1100 | 1.1568 | 48.0693 | 20.3882 |
85
+ | 0.1135 | 0.2 | 1200 | 1.1578 | 47.6190 | 20.0041 |
86
+ | 0.1055 | 0.2167 | 1300 | 1.1738 | 48.5541 | 19.8428 |
87
+ | 0.0973 | 0.2333 | 1400 | 1.1810 | 45.6277 | 18.7198 |
88
+ | 0.0941 | 0.25 | 1500 | 1.1755 | 47.0476 | 19.7666 |
89
+ | 0.073 | 0.2667 | 1600 | 1.1634 | 46.9437 | 19.3004 |
90
+ | 0.0797 | 0.2833 | 1700 | 1.1918 | 44.8658 | 18.3180 |
91
+ | 0.0623 | 0.3 | 1800 | 1.1941 | 43.2900 | 18.0424 |
92
+ | 0.061 | 0.3167 | 1900 | 1.1950 | 43.7229 | 18.5907 |
93
+ | 0.0568 | 0.3333 | 2000 | 1.1998 | 43.5498 | 17.3885 |
94
+ | 0.0516 | 0.35 | 2100 | 1.2007 | 44.2078 | 17.4999 |
95
+ | 0.0545 | 0.3667 | 2200 | 1.2149 | 48.1385 | 20.0715 |
96
+ | 0.0503 | 0.3833 | 2300 | 1.1990 | 46.9091 | 18.8576 |
97
+ | 0.0497 | 0.4 | 2400 | 1.1977 | 45.5238 | 18.5702 |
98
+ | 0.0497 | 0.4167 | 2500 | 1.2112 | 46.9610 | 18.7022 |
99
+ | 0.0341 | 0.4333 | 2600 | 1.2180 | 43.4286 | 17.1832 |
100
+ | 0.0497 | 0.45 | 2700 | 1.2105 | 45.9567 | 19.0247 |
101
+ | 0.0403 | 0.4667 | 2800 | 1.2209 | 43.6364 | 18.5380 |
102
+ | 0.0364 | 0.4833 | 2900 | 1.2249 | 43.0476 | 17.0161 |
103
+ | 0.0363 | 0.5 | 3000 | 1.2353 | 43.5671 | 17.3650 |
104
+ | 0.0366 | 0.5167 | 3100 | 1.2281 | 42.8052 | 16.9809 |
105
+ | 0.0278 | 0.5333 | 3200 | 1.2405 | 45.9221 | 18.6904 |
106
+ | 0.0334 | 0.55 | 3300 | 1.2316 | 46.3203 | 18.6377 |
107
+ | 0.0316 | 0.5667 | 3400 | 1.2474 | 46.6494 | 18.9368 |
108
+ | 0.0286 | 0.5833 | 3500 | 1.2504 | 49.5411 | 20.6903 |
109
+ | 0.0375 | 0.6 | 3600 | 1.2448 | 46.3203 | 18.3268 |
110
+ | 0.0228 | 0.6167 | 3700 | 1.2449 | 42.6494 | 16.9369 |
111
+ | 0.0197 | 0.6333 | 3800 | 1.2517 | 47.5152 | 19.4499 |
112
+ | 0.0185 | 1.0068 | 3900 | 1.2591 | 45.8009 | 18.5526 |
113
+ | 0.0245 | 1.0235 | 4000 | 1.2593 | 45.9913 | 18.7344 |
114
+ | 0.0146 | 1.0402 | 4100 | 1.2649 | 42.5281 | 18.1362 |
115
+ | 0.0138 | 1.0568 | 4200 | 1.2737 | 43.0130 | 16.9780 |
116
+ | 0.0245 | 1.0735 | 4300 | 1.2829 | 46.4762 | 18.9837 |
117
+ | 0.0271 | 1.0902 | 4400 | 1.2734 | 45.6797 | 18.4647 |
118
+ | 0.0237 | 1.1068 | 4500 | 1.2868 | 43.5498 | 17.3387 |
119
+ | 0.0211 | 1.1235 | 4600 | 1.2787 | 45.7143 | 18.6171 |
120
+ | 0.0366 | 1.1402 | 4700 | 1.2687 | 45.5931 | 18.4559 |
121
+ | 0.0279 | 1.1568 | 4800 | 1.2717 | 45.5584 | 18.3503 |
122
+ | 0.0233 | 1.1735 | 4900 | 1.2719 | 45.5584 | 18.4441 |
123
+ | 0.0264 | 1.1902 | 5000 | 1.2814 | 43.5844 | 17.4589 |
124
+ | 0.0245 | 1.2068 | 5100 | 1.2811 | 46.1126 | 18.4148 |
125
+ | 0.0162 | 1.2235 | 5200 | 1.2796 | 46.6667 | 19.1831 |
126
+ | 0.0204 | 1.2402 | 5300 | 1.2885 | 47.4286 | 20.5935 |
127
+ | 0.0214 | 1.2568 | 5400 | 1.2896 | 45.7489 | 18.4500 |
128
+ | 0.0163 | 1.2735 | 5500 | 1.2918 | 51.2900 | 22.1769 |
129
+ | 0.0151 | 1.2902 | 5600 | 1.2902 | 45.6104 | 18.4559 |
130
+ | 0.0195 | 1.3068 | 5700 | 1.2880 | 45.8874 | 18.8224 |
131
+ | 0.0134 | 1.3235 | 5800 | 1.2879 | 45.5584 | 18.5409 |
132
+ | 0.0137 | 1.3402 | 5900 | 1.2860 | 45.8528 | 18.8165 |
133
+ | 0.0174 | 1.3568 | 6000 | 1.2882 | 45.4199 | 18.4705 |
134
 
135
 
136
  ### Framework versions
 
139
  - Pytorch 2.3.0+cu121
140
  - Datasets 2.19.1
141
  - Tokenizers 0.19.1