Xlsr Waxal From Luganda

--- library_name: transformers license: apache-2.0 base_model: sulaimank/xls-r-300m-luganda-313hr tags: - generated_from_trainer metrics: - wer model-index: - name: xlsr-waxal-from-luganda results: [] --- # xlsr-waxal-from-luganda This model is a fine-tuned version of [sulaimank/xls-r-300m-luganda-313hr](https://huggingface.co/sulaimank/xls-r-300m-luganda-313hr) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.2325 - Wer: 0.2806 - Cer: 0.0853 - Wer Lug: 0.1613 - Wer Lin: 0.3334 - Wer Sna: 0.2719 ## 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: 0.0001 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 2 - total_train_batch_size: 32 - 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_steps: 0.1 - num_epochs: 30 ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | Cer | Wer Lug | Wer Lin | Wer Sna | |:-------------:|:-------:|:-----:|:---------------:|:------:|:------:|:-------:|:-------:|:-------:| | 21.2074 | 0.1463 | 500 | 8.4380 | 1.0000 | 0.9992 | 1.0 | 1.0 | 1.0000 | | 5.7880 | 0.2927 | 1000 | 2.8031 | 1.0000 | 0.9992 | 1.0 | 1.0 | 1.0000 | | 4.5646 | 0.4390 | 1500 | 2.0687 | 1.0 | 0.8928 | 1.0 | 1.0 | 1.0 | | 2.3249 | 0.5854 | 2000 | 1.0506 | 0.9775 | 0.3500 | 0.9771 | 0.9723 | 0.9841 | | 1.9122 | 0.7317 | 2500 | 0.8572 | 0.8284 | 0.2515 | 0.8643 | 0.8034 | 0.8425 | | 1.7780 | 0.8781 | 3000 | 0.7866 | 0.7892 | 0.2304 | 0.8237 | 0.7575 | 0.8126 | | 1.6314 | 1.0243 | 3500 | 0.7187 | 0.7340 | 0.2150 | 0.7808 | 0.7035 | 0.7497 | | 1.5220 | 1.1706 | 4000 | 0.6687 | 0.7017 | 0.2061 | 0.7197 | 0.6874 | 0.7109 | | 1.3840 | 1.3170 | 4500 | 0.6266 | 0.6613 | 0.1929 | 0.6246 | 0.6616 | 0.6790 | | 1.3338 | 1.4633 | 5000 | 0.6011 | 0.6584 | 0.1894 | 0.5561 | 0.6706 | 0.6929 | | 1.2607 | 1.6097 | 5500 | 0.5688 | 0.6123 | 0.1841 | 0.4791 | 0.6610 | 0.6157 | | 1.1818 | 1.7560 | 6000 | 0.5489 | 0.5583 | 0.1650 | 0.3907 | 0.5975 | 0.5903 | | 1.1072 | 1.9024 | 6500 | 0.5219 | 0.5283 | 0.1603 | 0.3396 | 0.5828 | 0.5514 | | 1.0281 | 2.0486 | 7000 | 0.5269 | 0.4928 | 0.1491 | 0.2937 | 0.5435 | 0.5258 | | 1.0194 | 2.1949 | 7500 | 0.4973 | 0.4780 | 0.1460 | 0.2756 | 0.5352 | 0.5044 | | 0.9406 | 2.3413 | 8000 | 0.4809 | 0.4662 | 0.1422 | 0.2665 | 0.5229 | 0.4919 | | 0.9777 | 2.4876 | 8500 | 0.4554 | 0.4772 | 0.1497 | 0.2523 | 0.5599 | 0.4821 | | 0.9098 | 2.6340 | 9000 | 0.4520 | 0.4407 | 0.1362 | 0.2392 | 0.5040 | 0.4589 | | 0.8793 | 2.7803 | 9500 | 0.4480 | 0.4909 | 0.1595 | 0.2357 | 0.6194 | 0.4527 | | 0.8249 | 2.9267 | 10000 | 0.4528 | 0.4148 | 0.1293 | 0.2216 | 0.4748 | 0.4332 | | 0.7515 | 3.0729 | 10500 | 0.4161 | 0.4199 | 0.1329 | 0.2142 | 0.4938 | 0.4268 | | 0.8012 | 3.2192 | 11000 | 0.4059 | 0.4221 | 0.1373 | 0.2116 | 0.5111 | 0.4120 | | 0.8088 | 3.3656 | 11500 | 0.4139 | 0.3970 | 0.1247 | 0.2044 | 0.4589 | 0.4127 | | 0.7414 | 3.5119 | 12000 | 0.4014 | 0.3875 | 0.1226 | 0.2003 | 0.4514 | 0.3979 | | 0.7193 | 3.6583 | 12500 | 0.3894 | 0.3810 | 0.1211 | 0.1988 | 0.4444 | 0.3897 | | 0.7677 | 3.8046 | 13000 | 0.3786 | 0.3926 | 0.1233 | 0.2030 | 0.4638 | 0.3948 | | 0.7347 | 3.9510 | 13500 | 0.3773 | 0.3737 | 0.1188 | 0.1946 | 0.4387 | 0.3787 | | 0.6657 | 4.0972 | 14000 | 0.3754 | 0.3722 | 0.1185 | 0.1924 | 0.4383 | 0.3762 | | 0.6621 | 4.2435 | 14500 | 0.3707 | 0.3641 | 0.1178 | 0.1940 | 0.4281 | 0.3661 | | 0.6837 | 4.3899 | 15000 | 0.3579 | 0.3647 | 0.1171 | 0.1835 | 0.4354 | 0.3635 | | 0.6827 | 4.5362 | 15500 | 0.3530 | 0.3703 | 0.1177 | 0.1903 | 0.4412 | 0.3682 | | 0.6236 | 4.6826 | 16000 | 0.3563 | 0.3800 | 0.1242 | 0.1909 | 0.4705 | 0.3575 | | 0.6701 | 4.8289 | 16500 | 0.3660 | 0.3530 | 0.1140 | 0.1827 | 0.4158 | 0.3566 | | 0.6811 | 4.9753 | 17000 | 0.3578 | 0.3484 | 0.1122 | 0.1814 | 0.4125 | 0.3486 | | 0.6206 | 5.1215 | 17500 | 0.3680 | 0.3467 | 0.1123 | 0.1829 | 0.4114 | 0.3447 | | 0.5978 | 5.2678 | 18000 | 0.3374 | 0.3520 | 0.1129 | 0.1837 | 0.4205 | 0.3474 | | 0.5872 | 5.4142 | 18500 | 0.3491 | 0.3442 | 0.1101 | 0.1821 | 0.4082 | 0.3421 | | 0.5911 | 5.5605 | 19000 | 0.3342 | 0.3503 | 0.1112 | 0.1800 | 0.4212 | 0.3437 | | 0.5771 | 5.7069 | 19500 | 0.3345 | 0.3519 | 0.1136 | 0.1800 | 0.4285 | 0.3389 | | 0.5751 | 5.8532 | 20000 | 0.3432 | 0.3374 | 0.1087 | 0.1768 | 0.4012 | 0.3350 | | 0.6158 | 5.9996 | 20500 | 0.3375 | 0.3353 | 0.1081 | 0.1761 | 0.3998 | 0.3314 | | 0.5215 | 6.1458 | 21000 | 0.3253 | 0.3431 | 0.1118 | 0.1787 | 0.4157 | 0.3313 | | 0.5423 | 6.2921 | 21500 | 0.3290 | 0.3350 | 0.1062 | 0.1779 | 0.3963 | 0.3341 | | 0.5467 | 6.4385 | 22000 | 0.3393 | 0.3308 | 0.1068 | 0.1744 | 0.3943 | 0.3266 | | 0.5765 | 6.5848 | 22500 | 0.3156 | 0.3320 | 0.1059 | 0.1739 | 0.3981 | 0.3254 | | 0.4922 | 6.7312 | 23000 | 0.3184 | 0.3273 | 0.1039 | 0.1757 | 0.3908 | 0.3208 | | 0.5498 | 6.8775 | 23500 | 0.3139 | 0.3290 | 0.1051 | 0.1747 | 0.3931 | 0.3231 | | 0.4997 | 7.0237 | 24000 | 0.3232 | 0.3242 | 0.1023 | 0.1731 | 0.3860 | 0.3196 | | 0.4635 | 7.1701 | 24500 | 0.3084 | 0.3283 | 0.1037 | 0.1717 | 0.3930 | 0.3229 | | 0.4957 | 7.3164 | 25000 | 0.3140 | 0.3239 | 0.1027 | 0.1720 | 0.3850 | 0.3208 | | 0.4840 | 7.4628 | 25500 | 0.3379 | 0.3215 | 0.1032 | 0.1734 | 0.3820 | 0.3171 | | 0.4712 | 7.6091 | 26000 | 0.3096 | 0.3225 | 0.1017 | 0.1741 | 0.3818 | 0.3199 | | 0.5330 | 7.7555 | 26500 | 0.3120 | 0.3200 | 0.1011 | 0.1717 | 0.3801 | 0.3163 | | 0.4769 | 7.9018 | 27000 | 0.2984 | 0.3210 | 0.1016 | 0.1719 | 0.3831 | 0.3153 | | 0.4623 | 8.0480 | 27500 | 0.3155 | 0.3156 | 0.0992 | 0.1701 | 0.3759 | 0.3102 | | 0.4614 | 8.1944 | 28000 | 0.3047 | 0.3166 | 0.1000 | 0.1703 | 0.3754 | 0.3136 | | 0.5024 | 8.3407 | 28500 | 0.2973 | 0.3154 | 0.0997 | 0.1684 | 0.3758 | 0.3106 | | 0.4791 | 8.4870 | 29000 | 0.2874 | 0.3153 | 0.0996 | 0.1688 | 0.3781 | 0.3072 | | 0.4726 | 8.6334 | 29500 | 0.2947 | 0.3137 | 0.0994 | 0.1702 | 0.3725 | 0.3092 | | 0.4847 | 8.7797 | 30000 | 0.2870 | 0.3161 | 0.0994 | 0.1685 | 0.3774 | 0.3105 | | 0.4695 | 8.9261 | 30500 | 0.2929 | 0.3111 | 0.0991 | 0.1699 | 0.3705 | 0.3046 | | 0.4348 | 9.0723 | 31000 | 0.2885 | 0.3100 | 0.0983 | 0.1705 | 0.3686 | 0.3039 | | 0.4265 | 9.2186 | 31500 | 0.2935 | 0.3095 | 0.0980 | 0.1689 | 0.3676 | 0.3045 | | 0.4537 | 9.3650 | 32000 | 0.2793 | 0.3099 | 0.0966 | 0.1675 | 0.3704 | 0.3026 | | 0.4559 | 9.5113 | 32500 | 0.2812 | 0.3091 | 0.0960 | 0.1683 | 0.3688 | 0.3024 | | 0.4180 | 9.6577 | 33000 | 0.2789 | 0.3100 | 0.0973 | 0.1672 | 0.3677 | 0.3067 | | 0.4263 | 9.8040 | 33500 | 0.2798 | 0.3067 | 0.0963 | 0.1689 | 0.3656 | 0.2992 | | 0.4584 | 9.9504 | 34000 | 0.2822 | 0.3054 | 0.0967 | 0.1689 | 0.3641 | 0.2978 | | 0.4005 | 10.0966 | 34500 | 0.2770 | 0.3058 | 0.0955 | 0.1691 | 0.3635 | 0.2995 | | 0.4018 | 10.2429 | 35000 | 0.2740 | 0.3071 | 0.0956 | 0.1663 | 0.3667 | 0.3004 | | 0.4474 | 10.3893 | 35500 | 0.2768 | 0.3039 | 0.0955 | 0.1676 | 0.3617 | 0.2971 | | 0.3940 | 10.5356 | 36000 | 0.2767 | 0.3025 | 0.0952 | 0.1657 | 0.3604 | 0.2958 | | 0.4062 | 10.6820 | 36500 | 0.2699 | 0.3025 | 0.0950 | 0.1672 | 0.3618 | 0.2935 | | 0.4176 | 10.8283 | 37000 | 0.2801 | 0.3006 | 0.0940 | 0.1656 | 0.3567 | 0.2953 | | 0.4230 | 10.9747 | 37500 | 0.2691 | 0.3005 | 0.0933 | 0.1659 | 0.3578 | 0.2936 | | 0.3564 | 11.1209 | 38000 | 0.2719 | 0.3011 | 0.0932 | 0.1649 | 0.3589 | 0.2943 | | 0.3958 | 11.2672 | 38500 | 0.2666 | 0.3006 | 0.0932 | 0.1658 | 0.3570 | 0.2950 | | 0.4042 | 11.4136 | 39000 | 0.2710 | 0.2985 | 0.0927 | 0.1653 | 0.3555 | 0.2912 | | 0.4263 | 11.5599 | 39500 | 0.2654 | 0.2976 | 0.0925 | 0.1639 | 0.3553 | 0.2897 | | 0.3904 | 11.7063 | 40000 | 0.2669 | 0.2962 | 0.0921 | 0.1629 | 0.3533 | 0.2889 | | 0.4115 | 11.8526 | 40500 | 0.2711 | 0.2963 | 0.0922 | 0.1650 | 0.3537 | 0.2877 | | 0.4133 | 11.9990 | 41000 | 0.2633 | 0.2966 | 0.0920 | 0.1633 | 0.3552 | 0.2875 | | 0.3925 | 12.1452 | 41500 | 0.2592 | 0.2967 | 0.0909 | 0.1637 | 0.3523 | 0.2913 | | 0.3853 | 12.2915 | 42000 | 0.2590 | 0.2995 | 0.0927 | 0.1657 | 0.3572 | 0.2917 | | 0.3938 | 12.4379 | 42500 | 0.2578 | 0.2941 | 0.0914 | 0.1634 | 0.3494 | 0.2879 | | 0.3861 | 12.5842 | 43000 | 0.2623 | 0.2926 | 0.0903 | 0.1640 | 0.3487 | 0.2842 | | 0.3716 | 12.7306 | 43500 | 0.2610 | 0.2919 | 0.0902 | 0.1626 | 0.3481 | 0.2839 | | 0.3574 | 12.8769 | 44000 | 0.2571 | 0.2933 | 0.0897 | 0.1621 | 0.3494 | 0.2862 | | 0.3830 | 13.0231 | 44500 | 0.2535 | 0.2938 | 0.0906 | 0.1620 | 0.3517 | 0.2848 | | 0.3598 | 13.1695 | 45000 | 0.2549 | 0.2938 | 0.0903 | 0.1641 | 0.3500 | 0.2859 | | 0.3697 | 13.3158 | 45500 | 0.2570 | 0.2912 | 0.0900 | 0.1626 | 0.3486 | 0.2813 | | 0.3275 | 13.4622 | 46000 | 0.2554 | 0.2910 | 0.0905 | 0.1621 | 0.3491 | 0.2804 | | 0.3743 | 13.6085 | 46500 | 0.2572 | 0.2898 | 0.0902 | 0.1625 | 0.3477 | 0.2785 | | 0.3529 | 13.7549 | 47000 | 0.2475 | 0.2914 | 0.0903 | 0.1622 | 0.3469 | 0.2842 | | 0.3496 | 13.9012 | 47500 | 0.2507 | 0.2900 | 0.0886 | 0.1624 | 0.3450 | 0.2827 | | 0.3649 | 14.0474 | 48000 | 0.2449 | 0.2920 | 0.0893 | 0.1616 | 0.3488 | 0.2838 | | 0.3395 | 14.1938 | 48500 | 0.2515 | 0.2882 | 0.0883 | 0.1597 | 0.3441 | 0.2801 | | 0.3236 | 14.3401 | 49000 | 0.2532 | 0.2866 | 0.0876 | 0.1609 | 0.3418 | 0.2780 | | 0.3521 | 14.4865 | 49500 | 0.2524 | 0.2852 | 0.0885 | 0.1599 | 0.3409 | 0.2756 | | 0.3577 | 14.6328 | 50000 | 0.2498 | 0.2850 | 0.0879 | 0.1608 | 0.3396 | 0.2764 | | 0.3328 | 14.7792 | 50500 | 0.2458 | 0.2849 | 0.0886 | 0.1607 | 0.3410 | 0.2744 | | 0.3523 | 14.9255 | 51000 | 0.2515 | 0.2836 | 0.0876 | 0.1594 | 0.3389 | 0.2743 | | 0.3218 | 15.0717 | 51500 | 0.2416 | 0.2842 | 0.0869 | 0.1598 | 0.3404 | 0.2736 | | 0.3319 | 15.2181 | 52000 | 0.2454 | 0.2849 | 0.0866 | 0.1621 | 0.3392 | 0.2759 | | 0.3337 | 15.3644 | 52500 | 0.2391 | 0.2838 | 0.0877 | 0.1610 | 0.3393 | 0.2735 | | 0.3001 | 15.5108 | 53000 | 0.2428 | 0.2827 | 0.0874 | 0.1600 | 0.3378 | 0.2727 | | 0.3059 | 15.6571 | 53500 | 0.2426 | 0.2832 | 0.0876 | 0.1587 | 0.3372 | 0.2755 | | 0.3347 | 15.8035 | 54000 | 0.2347 | 0.2866 | 0.0883 | 0.1586 | 0.3444 | 0.2758 | | 0.3460 | 15.9498 | 54500 | 0.2383 | 0.2832 | 0.0871 | 0.1587 | 0.3395 | 0.2726 | | 0.3352 | 16.0960 | 55000 | 0.2427 | 0.2803 | 0.0853 | 0.1569 | 0.3347 | 0.2715 | | 0.3435 | 16.2424 | 55500 | 0.2371 | 0.2835 | 0.0872 | 0.1584 | 0.3397 | 0.2732 | | 0.3291 | 16.3887 | 56000 | 0.2438 | 0.2796 | 0.0863 | 0.1593 | 0.3326 | 0.2711 | | 0.3042 | 16.5351 | 56500 | 0.2351 | 0.2790 | 0.0871 | 0.1595 | 0.3337 | 0.2681 | | 0.3092 | 16.6814 | 57000 | 0.2383 | 0.2790 | 0.0851 | 0.1585 | 0.3335 | 0.2688 | | 0.2947 | 16.8277 | 57500 | 0.2363 | 0.2793 | 0.0855 | 0.1598 | 0.3318 | 0.2711 | | 0.2999 | 16.9741 | 58000 | 0.2338 | 0.2800 | 0.0847 | 0.1601 | 0.3327 | 0.2717 | | 0.2970 | 17.1203 | 58500 | 0.2361 | 0.2797 | 0.0860 | 0.1616 | 0.3333 | 0.2695 | | 0.3251 | 17.2666 | 59000 | 0.2325 | 0.2806 | 0.0853 | 0.1613 | 0.3334 | 0.2719 | ### Framework versions - Transformers 5.12.1 - Pytorch 2.6.0+cu124 - Datasets 3.6.0 - Tokenizers 0.22.2