Whisper-Tiny-Java-v4
This model is a fine-tuned version of openai/whisper-tiny on the jv_id_asr_split dataset. It achieves the following results on the evaluation set:
- Loss: 0.1877
- Wer: 0.1560
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- training_steps: 50000
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
1.1788 | 0.0540 | 500 | 0.9671 | 0.6590 |
0.8015 | 0.1081 | 1000 | 0.6977 | 0.5305 |
0.6498 | 0.1621 | 1500 | 0.5725 | 0.6670 |
0.5828 | 0.2161 | 2000 | 0.5094 | 0.4829 |
0.5226 | 0.2702 | 2500 | 0.4642 | 0.3860 |
0.4955 | 0.3242 | 3000 | 0.4341 | 0.3915 |
0.4616 | 0.3782 | 3500 | 0.4128 | 0.3540 |
0.4474 | 0.4323 | 4000 | 0.3900 | 0.3614 |
0.4387 | 0.4863 | 4500 | 0.3736 | 0.3563 |
0.4154 | 0.5403 | 5000 | 0.3606 | 0.3274 |
0.419 | 0.5944 | 5500 | 0.3495 | 0.3144 |
0.3799 | 0.6484 | 6000 | 0.3398 | 0.2922 |
0.3802 | 0.7024 | 6500 | 0.3290 | 0.3044 |
0.3611 | 0.7565 | 7000 | 0.3225 | 0.2823 |
0.3548 | 0.8105 | 7500 | 0.3168 | 0.2733 |
0.346 | 0.8645 | 8000 | 0.3105 | 0.2660 |
0.3547 | 0.9186 | 8500 | 0.3063 | 0.2708 |
0.3211 | 0.9726 | 9000 | 0.3019 | 0.2827 |
0.2718 | 1.0267 | 9500 | 0.2990 | 0.2660 |
0.2859 | 1.0807 | 10000 | 0.2980 | 0.2587 |
0.2917 | 1.1348 | 10500 | 0.3269 | 0.2519 |
0.3117 | 1.1888 | 11000 | 0.3214 | 0.2575 |
0.3204 | 1.2428 | 11500 | 0.3168 | 0.2646 |
0.2962 | 1.2969 | 12000 | 0.3087 | 0.2410 |
0.2961 | 1.3509 | 12500 | 0.3057 | 0.2385 |
0.2887 | 1.4049 | 13000 | 0.2987 | 0.2281 |
0.2981 | 1.4590 | 13500 | 0.2953 | 0.2322 |
0.2994 | 1.5130 | 14000 | 0.2909 | 0.2322 |
0.2818 | 1.5670 | 14500 | 0.2848 | 0.2200 |
0.2851 | 1.6211 | 15000 | 0.2830 | 0.2166 |
0.275 | 1.6751 | 15500 | 0.2770 | 0.2129 |
0.2689 | 1.7291 | 16000 | 0.2760 | 0.2119 |
0.2796 | 1.7832 | 16500 | 0.2678 | 0.2002 |
0.2717 | 1.8372 | 17000 | 0.2653 | 0.2001 |
0.2661 | 1.8912 | 17500 | 0.2626 | 0.2014 |
0.2612 | 1.9453 | 18000 | 0.2573 | 0.1953 |
0.2532 | 1.9993 | 18500 | 0.2554 | 0.1954 |
0.1993 | 2.0534 | 19000 | 0.2527 | 0.1949 |
0.2009 | 2.1074 | 19500 | 0.2505 | 0.1897 |
0.1929 | 2.1615 | 20000 | 0.2484 | 0.1927 |
0.2011 | 2.2155 | 20500 | 0.2454 | 0.1895 |
0.1828 | 2.2695 | 21000 | 0.2444 | 0.1892 |
0.1823 | 2.3236 | 21500 | 0.2437 | 0.1845 |
0.186 | 2.3776 | 22000 | 0.2407 | 0.1839 |
0.1898 | 2.4316 | 22500 | 0.2390 | 0.1828 |
0.1789 | 2.4857 | 23000 | 0.2363 | 0.1790 |
0.1765 | 2.5397 | 23500 | 0.2353 | 0.1797 |
0.1808 | 2.5937 | 24000 | 0.2320 | 0.1797 |
0.1771 | 2.6478 | 24500 | 0.2291 | 0.1777 |
0.183 | 2.7018 | 25000 | 0.2276 | 0.1788 |
0.178 | 2.7558 | 25500 | 0.2250 | 0.1754 |
0.1829 | 2.8099 | 26000 | 0.2231 | 0.1755 |
0.183 | 2.8639 | 26500 | 0.2216 | 0.1790 |
0.1812 | 2.9179 | 27000 | 0.2198 | 0.1729 |
0.1697 | 2.9720 | 27500 | 0.2186 | 0.1727 |
0.1317 | 3.0260 | 28000 | 0.2173 | 0.1728 |
0.1298 | 3.0801 | 28500 | 0.2159 | 0.1690 |
0.1272 | 3.1341 | 29000 | 0.2161 | 0.1686 |
0.1389 | 3.1881 | 29500 | 0.2148 | 0.1706 |
0.1379 | 3.2422 | 30000 | 0.2139 | 0.1693 |
0.1312 | 3.2962 | 30500 | 0.2133 | 0.1714 |
0.1212 | 3.3502 | 31000 | 0.2116 | 0.1706 |
0.1265 | 3.4043 | 31500 | 0.2103 | 0.1666 |
0.1261 | 3.4583 | 32000 | 0.2095 | 0.1706 |
0.127 | 3.5123 | 32500 | 0.2079 | 0.1673 |
0.1346 | 3.5664 | 33000 | 0.2061 | 0.1683 |
0.1283 | 3.6204 | 33500 | 0.2046 | 0.1652 |
0.1244 | 3.6744 | 34000 | 0.2040 | 0.1684 |
0.1207 | 3.7285 | 34500 | 0.2026 | 0.1648 |
0.1239 | 3.7825 | 35000 | 0.2022 | 0.1622 |
0.1308 | 3.8365 | 35500 | 0.1998 | 0.1624 |
0.1272 | 3.8906 | 36000 | 0.1997 | 0.1649 |
0.1328 | 3.9446 | 36500 | 0.1988 | 0.1647 |
0.1256 | 3.9986 | 37000 | 0.1971 | 0.1653 |
0.0953 | 4.0527 | 37500 | 0.1974 | 0.1604 |
0.0946 | 4.1068 | 38000 | 0.1979 | 0.1625 |
0.0933 | 4.1608 | 38500 | 0.1964 | 0.1609 |
0.1025 | 4.2148 | 39000 | 0.1962 | 0.1634 |
0.1002 | 4.2689 | 39500 | 0.1957 | 0.1632 |
0.0976 | 4.3229 | 40000 | 0.1949 | 0.1621 |
0.0983 | 4.3769 | 40500 | 0.1936 | 0.1605 |
0.0995 | 4.4310 | 41000 | 0.1935 | 0.1608 |
0.0877 | 4.4850 | 41500 | 0.1930 | 0.1608 |
0.0985 | 4.5390 | 42000 | 0.1928 | 0.1633 |
0.0887 | 4.5931 | 42500 | 0.1917 | 0.1616 |
0.0909 | 4.6471 | 43000 | 0.1918 | 0.1604 |
0.0908 | 4.7011 | 43500 | 0.1910 | 0.1593 |
0.0931 | 4.7552 | 44000 | 0.1902 | 0.1579 |
0.0938 | 4.8092 | 44500 | 0.1890 | 0.1582 |
0.0925 | 4.8632 | 45000 | 0.1889 | 0.1594 |
0.0943 | 4.9173 | 45500 | 0.1882 | 0.1578 |
0.0918 | 4.9713 | 46000 | 0.1879 | 0.1584 |
0.0791 | 5.0253 | 46500 | 0.1877 | 0.1560 |
0.077 | 5.0793 | 47000 | 0.1877 | 0.1587 |
0.0769 | 5.1334 | 47500 | 0.1878 | 0.1597 |
0.0744 | 5.1874 | 48000 | 0.1876 | 0.1585 |
0.0775 | 5.2414 | 48500 | 0.1874 | 0.1595 |
0.069 | 5.2955 | 49000 | 0.1873 | 0.1579 |
0.0761 | 5.3495 | 49500 | 0.1870 | 0.1575 |
0.0711 | 5.4036 | 50000 | 0.1869 | 0.1583 |
Framework versions
- Transformers 4.50.0.dev0
- Pytorch 2.6.0+cu126
- Datasets 2.16.0
- Tokenizers 0.21.1
- Downloads last month
- 6
Model tree for bagasshw/whisper-tiny-javanese-openslr-v4
Base model
openai/whisper-tiny