--- library_name: transformers language: - ja license: apache-2.0 base_model: rinna/japanese-hubert-base tags: - automatic-speech-recognition - mozilla-foundation/common_voice_13_0 - generated_from_trainer datasets: - common_voice_13_0 metrics: - wer model-index: - name: Hubert-common_voice-ja-demo-kana-debug-50epochs-cosine results: - task: name: Automatic Speech Recognition type: automatic-speech-recognition dataset: name: MOZILLA-FOUNDATION/COMMON_VOICE_13_0 - JA type: common_voice_13_0 config: ja split: test args: 'Config: ja, Training split: train+validation, Eval split: test' metrics: - name: Wer type: wer value: 1.0 --- # Hubert-common_voice-ja-demo-kana-debug-50epochs-cosine This model is a fine-tuned version of [rinna/japanese-hubert-base](https://huggingface.co/rinna/japanese-hubert-base) on the MOZILLA-FOUNDATION/COMMON_VOICE_13_0 - JA dataset. It achieves the following results on the evaluation set: - Loss: 0.6997 - Wer: 1.0 - Cer: 0.3178 ## 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: 3e-05 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 2 - total_train_batch_size: 32 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments - lr_scheduler_type: cosine - lr_scheduler_warmup_steps: 12500 - num_epochs: 50.0 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | Cer | |:-------------:|:-------:|:-----:|:---------------:|:------:|:------:| | No log | 0.2660 | 100 | 43.2254 | 1.5295 | 5.8209 | | No log | 0.5319 | 200 | 42.4473 | 1.5328 | 5.2847 | | No log | 0.7979 | 300 | 40.4745 | 1.1351 | 1.8916 | | No log | 1.0638 | 400 | 33.1315 | 1.0 | 0.9999 | | 31.9649 | 1.3298 | 500 | 20.9238 | 1.0 | 0.9999 | | 31.9649 | 1.5957 | 600 | 17.6015 | 1.0 | 0.9999 | | 31.9649 | 1.8617 | 700 | 16.8689 | 1.0 | 0.9999 | | 31.9649 | 2.1277 | 800 | 16.4459 | 1.0 | 0.9999 | | 31.9649 | 2.3936 | 900 | 16.0021 | 1.0 | 0.9999 | | 14.4801 | 2.6596 | 1000 | 15.5383 | 1.0 | 0.9999 | | 14.4801 | 2.9255 | 1100 | 15.0419 | 1.0 | 0.9999 | | 14.4801 | 3.1915 | 1200 | 14.5001 | 1.0 | 0.9999 | | 14.4801 | 3.4574 | 1300 | 13.9273 | 1.0 | 0.9999 | | 14.4801 | 3.7234 | 1400 | 13.3177 | 1.0 | 0.9999 | | 12.1603 | 3.9894 | 1500 | 12.6793 | 1.0 | 0.9999 | | 12.1603 | 4.2553 | 1600 | 12.0061 | 1.0 | 0.9999 | | 12.1603 | 4.5213 | 1700 | 11.3136 | 1.0 | 0.9999 | | 12.1603 | 4.7872 | 1800 | 10.5996 | 1.0 | 0.9999 | | 12.1603 | 5.0532 | 1900 | 9.8780 | 1.0 | 0.9999 | | 9.51 | 5.3191 | 2000 | 9.1535 | 1.0 | 0.9999 | | 9.51 | 5.5851 | 2100 | 8.4424 | 1.0 | 0.9999 | | 9.51 | 5.8511 | 2200 | 7.7545 | 1.0 | 0.9999 | | 9.51 | 6.1170 | 2300 | 7.1014 | 1.0 | 0.9999 | | 9.51 | 6.3830 | 2400 | 6.4994 | 1.0 | 0.9999 | | 6.6851 | 6.6489 | 2500 | 5.9542 | 1.0 | 0.9999 | | 6.6851 | 6.9149 | 2600 | 5.4843 | 1.0 | 0.9999 | | 6.6851 | 7.1809 | 2700 | 5.0985 | 1.0 | 0.9999 | | 6.6851 | 7.4468 | 2800 | 4.7872 | 1.0 | 0.9999 | | 6.6851 | 7.7128 | 2900 | 4.5565 | 1.0 | 0.9999 | | 4.7438 | 7.9787 | 3000 | 4.3885 | 1.0 | 0.9999 | | 4.7438 | 8.2447 | 3100 | 4.2609 | 1.0 | 0.9999 | | 4.7438 | 8.5106 | 3200 | 4.1909 | 1.0 | 0.9999 | | 4.7438 | 8.7766 | 3300 | 4.1393 | 1.0 | 0.9999 | | 4.7438 | 9.0426 | 3400 | 4.1096 | 1.0 | 0.9999 | | 4.1247 | 9.3085 | 3500 | 4.0906 | 1.0 | 0.9999 | | 4.1247 | 9.5745 | 3600 | 4.0835 | 1.0 | 0.9999 | | 4.1247 | 9.8404 | 3700 | 4.0710 | 1.0 | 0.9999 | | 4.1247 | 10.1064 | 3800 | 4.0627 | 1.0 | 0.9999 | | 4.1247 | 10.3723 | 3900 | 4.0562 | 1.0 | 0.9999 | | 4.039 | 10.6383 | 4000 | 4.0456 | 1.0 | 0.9999 | | 4.039 | 10.9043 | 4100 | 4.0376 | 1.0 | 0.9999 | | 4.039 | 11.1702 | 4200 | 4.0261 | 1.0 | 0.9999 | | 4.039 | 11.4362 | 4300 | 4.0129 | 1.0 | 0.9999 | | 4.039 | 11.7021 | 4400 | 4.0102 | 1.0 | 0.9999 | | 3.9934 | 11.9681 | 4500 | 3.9765 | 1.0 | 0.9999 | | 3.9934 | 12.2340 | 4600 | 3.9292 | 1.0 | 0.9999 | | 3.9934 | 12.5 | 4700 | 3.8597 | 1.0 | 0.9999 | | 3.9934 | 12.7660 | 4800 | 3.7626 | 1.0 | 0.9999 | | 3.9934 | 13.0319 | 4900 | 3.6057 | 1.0 | 0.9999 | | 3.7495 | 13.2979 | 5000 | 3.4160 | 1.0 | 0.9999 | | 3.7495 | 13.5638 | 5100 | 3.1636 | 1.0 | 0.9998 | | 3.7495 | 13.8298 | 5200 | 2.7153 | 1.0 | 0.7268 | | 3.7495 | 14.0957 | 5300 | 2.4242 | 1.0 | 0.5731 | | 3.7495 | 14.3617 | 5400 | 2.1946 | 1.0 | 0.5367 | | 2.6537 | 14.6277 | 5500 | 2.0187 | 1.0 | 0.4742 | | 2.6537 | 14.8936 | 5600 | 1.8818 | 1.0 | 0.4641 | | 2.6537 | 15.1596 | 5700 | 1.7484 | 1.0 | 0.4355 | | 2.6537 | 15.4255 | 5800 | 1.6397 | 1.0 | 0.4217 | | 2.6537 | 15.6915 | 5900 | 1.5507 | 0.9998 | 0.4047 | | 1.7492 | 15.9574 | 6000 | 1.4586 | 1.0 | 0.3999 | | 1.7492 | 16.2234 | 6100 | 1.3908 | 1.0 | 0.3913 | | 1.7492 | 16.4894 | 6200 | 1.3182 | 1.0 | 0.3849 | | 1.7492 | 16.7553 | 6300 | 1.2754 | 1.0 | 0.3739 | | 1.7492 | 17.0213 | 6400 | 1.2096 | 1.0 | 0.3744 | | 1.3182 | 17.2872 | 6500 | 1.1599 | 1.0 | 0.3713 | | 1.3182 | 17.5532 | 6600 | 1.1126 | 1.0 | 0.3673 | | 1.3182 | 17.8191 | 6700 | 1.0685 | 1.0 | 0.3642 | | 1.3182 | 18.0851 | 6800 | 1.0331 | 1.0 | 0.3606 | | 1.3182 | 18.3511 | 6900 | 1.0006 | 1.0 | 0.3581 | | 1.06 | 18.6170 | 7000 | 0.9663 | 1.0 | 0.3570 | | 1.06 | 18.8830 | 7100 | 0.9409 | 1.0 | 0.3541 | | 1.06 | 19.1489 | 7200 | 0.9141 | 0.9998 | 0.3527 | | 1.06 | 19.4149 | 7300 | 0.8919 | 1.0 | 0.3515 | | 1.06 | 19.6809 | 7400 | 0.8824 | 0.9998 | 0.3516 | | 0.8957 | 19.9468 | 7500 | 0.8395 | 1.0 | 0.3477 | | 0.8957 | 20.2128 | 7600 | 0.8356 | 1.0 | 0.3476 | | 0.8957 | 20.4787 | 7700 | 0.8232 | 1.0 | 0.3506 | | 0.8957 | 20.7447 | 7800 | 0.7880 | 1.0 | 0.3443 | | 0.8957 | 21.0106 | 7900 | 0.7795 | 1.0 | 0.3453 | | 0.7596 | 21.2766 | 8000 | 0.7670 | 1.0 | 0.3435 | | 0.7596 | 21.5426 | 8100 | 0.7575 | 1.0 | 0.3426 | | 0.7596 | 21.8085 | 8200 | 0.7344 | 1.0 | 0.3415 | | 0.7596 | 22.0745 | 8300 | 0.7258 | 1.0 | 0.3413 | | 0.7596 | 22.3404 | 8400 | 0.7230 | 0.9998 | 0.3410 | | 0.6631 | 22.6064 | 8500 | 0.7223 | 1.0 | 0.3390 | | 0.6631 | 22.8723 | 8600 | 0.7035 | 0.9998 | 0.3391 | | 0.6631 | 23.1383 | 8700 | 0.7013 | 1.0 | 0.3370 | | 0.6631 | 23.4043 | 8800 | 0.6926 | 1.0 | 0.3365 | | 0.6631 | 23.6702 | 8900 | 0.6865 | 1.0 | 0.3368 | | 0.6032 | 23.9362 | 9000 | 0.6834 | 1.0 | 0.3372 | | 0.6032 | 24.2021 | 9100 | 0.6720 | 0.9998 | 0.3356 | | 0.6032 | 24.4681 | 9200 | 0.6614 | 1.0 | 0.3353 | | 0.6032 | 24.7340 | 9300 | 0.6691 | 1.0 | 0.3352 | | 0.6032 | 25.0 | 9400 | 0.6571 | 0.9998 | 0.3328 | | 0.544 | 25.2660 | 9500 | 0.6790 | 1.0 | 0.3360 | | 0.544 | 25.5319 | 9600 | 0.6571 | 1.0 | 0.3326 | | 0.544 | 25.7979 | 9700 | 0.6508 | 1.0 | 0.3344 | | 0.544 | 26.0638 | 9800 | 0.6482 | 1.0 | 0.3327 | | 0.544 | 26.3298 | 9900 | 0.6354 | 0.9998 | 0.3312 | | 0.4943 | 26.5957 | 10000 | 0.6280 | 0.9998 | 0.3308 | | 0.4943 | 26.8617 | 10100 | 0.6407 | 1.0 | 0.3299 | | 0.4943 | 27.1277 | 10200 | 0.6388 | 1.0 | 0.3311 | | 0.4943 | 27.3936 | 10300 | 0.6483 | 1.0 | 0.3302 | | 0.4943 | 27.6596 | 10400 | 0.6255 | 0.9998 | 0.3311 | | 0.4613 | 27.9255 | 10500 | 0.6355 | 1.0 | 0.3308 | | 0.4613 | 28.1915 | 10600 | 0.6293 | 1.0 | 0.3300 | | 0.4613 | 28.4574 | 10700 | 0.6333 | 1.0 | 0.3288 | | 0.4613 | 28.7234 | 10800 | 0.6207 | 1.0 | 0.3279 | | 0.4613 | 28.9894 | 10900 | 0.6218 | 0.9998 | 0.3281 | | 0.4151 | 29.2553 | 11000 | 0.6301 | 1.0 | 0.3295 | | 0.4151 | 29.5213 | 11100 | 0.6189 | 0.9996 | 0.3300 | | 0.4151 | 29.7872 | 11200 | 0.6250 | 0.9998 | 0.3279 | | 0.4151 | 30.0532 | 11300 | 0.6211 | 1.0 | 0.3268 | | 0.4151 | 30.3191 | 11400 | 0.6195 | 1.0 | 0.3278 | | 0.3895 | 30.5851 | 11500 | 0.6265 | 0.9998 | 0.3270 | | 0.3895 | 30.8511 | 11600 | 0.6332 | 1.0 | 0.3263 | | 0.3895 | 31.1170 | 11700 | 0.6232 | 0.9998 | 0.3253 | | 0.3895 | 31.3830 | 11800 | 0.6281 | 1.0 | 0.3262 | | 0.3895 | 31.6489 | 11900 | 0.6212 | 1.0 | 0.3258 | | 0.3686 | 31.9149 | 12000 | 0.6368 | 0.9998 | 0.3266 | | 0.3686 | 32.1809 | 12100 | 0.6276 | 0.9998 | 0.3299 | | 0.3686 | 32.4468 | 12200 | 0.6513 | 1.0 | 0.3333 | | 0.3686 | 32.7128 | 12300 | 0.6249 | 0.9996 | 0.3229 | | 0.3686 | 32.9787 | 12400 | 0.6232 | 1.0 | 0.3262 | | 0.3415 | 33.2447 | 12500 | 0.6144 | 1.0 | 0.3240 | | 0.3415 | 33.5106 | 12600 | 0.6243 | 1.0 | 0.3269 | | 0.3415 | 33.7766 | 12700 | 0.6344 | 1.0 | 0.3249 | | 0.3415 | 34.0426 | 12800 | 0.6372 | 1.0 | 0.3236 | | 0.3415 | 34.3085 | 12900 | 0.6399 | 1.0 | 0.3247 | | 0.3167 | 34.5745 | 13000 | 0.6329 | 0.9996 | 0.3231 | | 0.3167 | 34.8404 | 13100 | 0.6251 | 0.9998 | 0.3249 | | 0.3167 | 35.1064 | 13200 | 0.6508 | 0.9998 | 0.3234 | | 0.3167 | 35.3723 | 13300 | 0.6473 | 1.0 | 0.3219 | | 0.3167 | 35.6383 | 13400 | 0.7159 | 0.9998 | 0.3232 | | 0.3006 | 35.9043 | 13500 | 0.6520 | 0.9998 | 0.3279 | | 0.3006 | 36.1702 | 13600 | 0.6568 | 0.9998 | 0.3225 | | 0.3006 | 36.4362 | 13700 | 0.6568 | 1.0 | 0.3221 | | 0.3006 | 36.7021 | 13800 | 0.6531 | 0.9998 | 0.3211 | | 0.3006 | 36.9681 | 13900 | 0.6393 | 1.0 | 0.3218 | | 0.2769 | 37.2340 | 14000 | 0.6671 | 0.9996 | 0.3229 | | 0.2769 | 37.5 | 14100 | 0.6516 | 0.9998 | 0.3227 | | 0.2769 | 37.7660 | 14200 | 0.6632 | 1.0 | 0.3218 | | 0.2769 | 38.0319 | 14300 | 0.6799 | 1.0 | 0.3220 | | 0.2769 | 38.2979 | 14400 | 0.6666 | 1.0 | 0.3220 | | 0.2611 | 38.5638 | 14500 | 0.6486 | 1.0 | 0.3209 | | 0.2611 | 38.8298 | 14600 | 0.6458 | 1.0 | 0.3209 | | 0.2611 | 39.0957 | 14700 | 0.7051 | 1.0 | 0.3208 | | 0.2611 | 39.3617 | 14800 | 0.6990 | 0.9998 | 0.3217 | | 0.2611 | 39.6277 | 14900 | 0.6779 | 1.0 | 0.3216 | | 0.2399 | 39.8936 | 15000 | 0.6609 | 1.0 | 0.3215 | | 0.2399 | 40.1596 | 15100 | 0.6820 | 1.0 | 0.3208 | | 0.2399 | 40.4255 | 15200 | 0.6618 | 1.0 | 0.3202 | | 0.2399 | 40.6915 | 15300 | 0.6736 | 0.9998 | 0.3190 | | 0.2399 | 40.9574 | 15400 | 0.6720 | 1.0 | 0.3196 | | 0.2251 | 41.2234 | 15500 | 0.6697 | 1.0 | 0.3190 | | 0.2251 | 41.4894 | 15600 | 0.6754 | 1.0 | 0.3214 | | 0.2251 | 41.7553 | 15700 | 0.6777 | 1.0 | 0.3193 | | 0.2251 | 42.0213 | 15800 | 0.6920 | 1.0 | 0.3195 | | 0.2251 | 42.2872 | 15900 | 0.7009 | 1.0 | 0.3189 | | 0.2181 | 42.5532 | 16000 | 0.6936 | 1.0 | 0.3184 | | 0.2181 | 42.8191 | 16100 | 0.6941 | 1.0 | 0.3191 | | 0.2181 | 43.0851 | 16200 | 0.6952 | 1.0 | 0.3190 | | 0.2181 | 43.3511 | 16300 | 0.7078 | 1.0 | 0.3193 | | 0.2181 | 43.6170 | 16400 | 0.6847 | 0.9998 | 0.3183 | | 0.2019 | 43.8830 | 16500 | 0.6928 | 1.0 | 0.3203 | | 0.2019 | 44.1489 | 16600 | 0.6941 | 1.0 | 0.3189 | | 0.2019 | 44.4149 | 16700 | 0.6907 | 1.0 | 0.3180 | | 0.2019 | 44.6809 | 16800 | 0.6850 | 0.9998 | 0.3183 | | 0.2019 | 44.9468 | 16900 | 0.6956 | 1.0 | 0.3181 | | 0.1938 | 45.2128 | 17000 | 0.6916 | 0.9998 | 0.3178 | | 0.1938 | 45.4787 | 17100 | 0.6940 | 0.9996 | 0.3178 | | 0.1938 | 45.7447 | 17200 | 0.6952 | 0.9998 | 0.3190 | | 0.1938 | 46.0106 | 17300 | 0.7024 | 1.0 | 0.3189 | | 0.1938 | 46.2766 | 17400 | 0.7070 | 0.9998 | 0.3188 | | 0.1908 | 46.5426 | 17500 | 0.7046 | 0.9998 | 0.3182 | | 0.1908 | 46.8085 | 17600 | 0.7046 | 0.9996 | 0.3183 | | 0.1908 | 47.0745 | 17700 | 0.7030 | 1.0 | 0.3181 | | 0.1908 | 47.3404 | 17800 | 0.6982 | 0.9998 | 0.3181 | | 0.1908 | 47.6064 | 17900 | 0.7023 | 1.0 | 0.3177 | | 0.1805 | 47.8723 | 18000 | 0.6954 | 1.0 | 0.3180 | | 0.1805 | 48.1383 | 18100 | 0.6969 | 0.9998 | 0.3179 | | 0.1805 | 48.4043 | 18200 | 0.6992 | 0.9998 | 0.3177 | | 0.1805 | 48.6702 | 18300 | 0.6975 | 1.0 | 0.3176 | | 0.1805 | 48.9362 | 18400 | 0.6991 | 0.9998 | 0.3177 | | 0.1817 | 49.2021 | 18500 | 0.6993 | 0.9998 | 0.3175 | | 0.1817 | 49.4681 | 18600 | 0.7004 | 0.9998 | 0.3177 | | 0.1817 | 49.7340 | 18700 | 0.7000 | 1.0 | 0.3178 | | 0.1817 | 50.0 | 18800 | 0.7000 | 0.9996 | 0.3176 | ### Framework versions - Transformers 4.47.0.dev0 - Pytorch 2.5.1+cu124 - Datasets 3.1.0 - Tokenizers 0.20.3