--- license: apache-2.0 base_model: facebook/wav2vec2-large-xlsr-53 tags: - automatic-speech-recognition - DewiBrynJones/banc-trawsgrifiadau-bangor-clean-with-ccv - generated_from_trainer metrics: - wer model-index: - name: wav2vec2-xlsr-53-ft-btb-ccv-cy results: [] --- # wav2vec2-xlsr-53-ft-btb-ccv-cy This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the DEWIBRYNJONES/BANC-TRAWSGRIFIADAU-BANGOR-CLEAN-WITH-CCV - DEFAULT dataset. It achieves the following results on the evaluation set: - Loss: 0.4511 - Wer: 0.3591 ## 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: 0.0003 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - training_steps: 30000 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | |:-------------:|:------:|:-----:|:---------------:|:------:| | No log | 0.0079 | 200 | 3.1893 | 1.0 | | No log | 0.0157 | 400 | 2.7802 | 1.0 | | 4.719 | 0.0236 | 600 | 1.4221 | 0.8877 | | 4.719 | 0.0314 | 800 | 1.2274 | 0.8224 | | 1.0441 | 0.0393 | 1000 | 1.1095 | 0.7887 | | 1.0441 | 0.0472 | 1200 | 1.0914 | 0.7549 | | 1.0441 | 0.0550 | 1400 | 1.0177 | 0.7355 | | 0.8033 | 0.0629 | 1600 | 0.9907 | 0.7233 | | 0.8033 | 0.0707 | 1800 | 0.9761 | 0.7145 | | 0.7227 | 0.0786 | 2000 | 0.9555 | 0.6903 | | 0.7227 | 0.0864 | 2200 | 0.8995 | 0.6748 | | 0.7227 | 0.0943 | 2400 | 0.8897 | 0.6666 | | 0.6794 | 0.1022 | 2600 | 0.8826 | 0.6560 | | 0.6794 | 0.1100 | 2800 | 0.8745 | 0.6446 | | 0.6513 | 0.1179 | 3000 | 0.8450 | 0.6437 | | 0.6513 | 0.1257 | 3200 | 0.8596 | 0.6511 | | 0.6513 | 0.1336 | 3400 | 0.8598 | 0.6376 | | 0.6147 | 0.1415 | 3600 | 0.8516 | 0.6375 | | 0.6147 | 0.1493 | 3800 | 0.8252 | 0.6100 | | 0.6092 | 0.1572 | 4000 | 0.8580 | 0.6823 | | 0.6092 | 0.1650 | 4200 | 0.8205 | 0.6136 | | 0.6092 | 0.1729 | 4400 | 0.8033 | 0.6385 | | 0.5928 | 0.1808 | 4600 | 0.7928 | 0.6005 | | 0.5928 | 0.1886 | 4800 | 0.7911 | 0.5924 | | 0.5681 | 0.1965 | 5000 | 0.7969 | 0.5944 | | 0.5681 | 0.2043 | 5200 | 0.7933 | 0.5899 | | 0.5681 | 0.2122 | 5400 | 0.7830 | 0.6013 | | 0.5806 | 0.2200 | 5600 | 0.7703 | 0.5789 | | 0.5806 | 0.2279 | 5800 | 0.7666 | 0.5898 | | 0.5608 | 0.2358 | 6000 | 0.7580 | 0.5695 | | 0.5608 | 0.2436 | 6200 | 0.7479 | 0.5651 | | 0.5608 | 0.2515 | 6400 | 0.7639 | 0.5847 | | 0.5333 | 0.2593 | 6600 | 0.7297 | 0.5676 | | 0.5333 | 0.2672 | 6800 | 0.7441 | 0.5590 | | 0.5406 | 0.2751 | 7000 | 0.7405 | 0.5491 | | 0.5406 | 0.2829 | 7200 | 0.7238 | 0.5529 | | 0.5406 | 0.2908 | 7400 | 0.7328 | 0.5544 | | 0.535 | 0.2986 | 7600 | 0.7263 | 0.5599 | | 0.535 | 0.3065 | 7800 | 0.7421 | 0.5594 | | 0.5195 | 0.3144 | 8000 | 0.7435 | 0.5544 | | 0.5195 | 0.3222 | 8200 | 0.7187 | 0.5424 | | 0.5195 | 0.3301 | 8400 | 0.6977 | 0.5353 | | 0.5023 | 0.3379 | 8600 | 0.6950 | 0.5386 | | 0.5023 | 0.3458 | 8800 | 0.7155 | 0.5451 | | 0.5106 | 0.3536 | 9000 | 0.6857 | 0.5379 | | 0.5106 | 0.3615 | 9200 | 0.6848 | 0.5329 | | 0.5106 | 0.3694 | 9400 | 0.6732 | 0.5202 | | 0.4968 | 0.3772 | 9600 | 0.6839 | 0.5275 | | 0.4968 | 0.3851 | 9800 | 0.6767 | 0.5198 | | 0.4824 | 0.3929 | 10000 | 0.6718 | 0.5335 | | 0.4824 | 0.4008 | 10200 | 0.6593 | 0.5175 | | 0.4824 | 0.4087 | 10400 | 0.6799 | 0.5174 | | 0.48 | 0.4165 | 10600 | 0.6662 | 0.5129 | | 0.48 | 0.4244 | 10800 | 0.6619 | 0.5006 | | 0.4693 | 0.4322 | 11000 | 0.6576 | 0.5199 | | 0.4693 | 0.4401 | 11200 | 0.6406 | 0.5019 | | 0.4693 | 0.4480 | 11400 | 0.6408 | 0.5066 | | 0.4691 | 0.4558 | 11600 | 0.6476 | 0.5019 | | 0.4691 | 0.4637 | 11800 | 0.6423 | 0.4946 | | 0.4444 | 0.4715 | 12000 | 0.6374 | 0.4976 | | 0.4444 | 0.4794 | 12200 | 0.6312 | 0.4961 | | 0.4444 | 0.4872 | 12400 | 0.6170 | 0.4819 | | 0.4474 | 0.4951 | 12600 | 0.6301 | 0.4933 | | 0.4474 | 0.5030 | 12800 | 0.6253 | 0.4862 | | 0.4471 | 0.5108 | 13000 | 0.6220 | 0.4849 | | 0.4471 | 0.5187 | 13200 | 0.6201 | 0.4853 | | 0.4471 | 0.5265 | 13400 | 0.6168 | 0.4848 | | 0.4323 | 0.5344 | 13600 | 0.6173 | 0.4771 | | 0.4323 | 0.5423 | 13800 | 0.6032 | 0.4656 | | 0.4575 | 0.5501 | 14000 | 0.6097 | 0.4678 | | 0.4575 | 0.5580 | 14200 | 0.5971 | 0.4674 | | 0.4575 | 0.5658 | 14400 | 0.5977 | 0.4698 | | 0.4395 | 0.5737 | 14600 | 0.6057 | 0.4734 | | 0.4395 | 0.5816 | 14800 | 0.5827 | 0.4574 | | 0.4119 | 0.5894 | 15000 | 0.5946 | 0.4640 | | 0.4119 | 0.5973 | 15200 | 0.6023 | 0.4771 | | 0.4119 | 0.6051 | 15400 | 0.6129 | 0.4727 | | 0.4125 | 0.6130 | 15600 | 0.5902 | 0.4584 | | 0.4125 | 0.6208 | 15800 | 0.5955 | 0.4654 | | 0.4039 | 0.6287 | 16000 | 0.5955 | 0.4595 | | 0.4039 | 0.6366 | 16200 | 0.5789 | 0.4497 | | 0.4039 | 0.6444 | 16400 | 0.5779 | 0.4630 | | 0.3969 | 0.6523 | 16600 | 0.5677 | 0.4551 | | 0.3969 | 0.6601 | 16800 | 0.5869 | 0.4606 | | 0.3923 | 0.6680 | 17000 | 0.5710 | 0.4502 | | 0.3923 | 0.6759 | 17200 | 0.5640 | 0.4474 | | 0.3923 | 0.6837 | 17400 | 0.5842 | 0.4498 | | 0.386 | 0.6916 | 17600 | 0.5597 | 0.4440 | | 0.386 | 0.6994 | 17800 | 0.5621 | 0.4381 | | 0.3851 | 0.7073 | 18000 | 0.5665 | 0.4346 | | 0.3851 | 0.7152 | 18200 | 0.5573 | 0.4356 | | 0.3851 | 0.7230 | 18400 | 0.5548 | 0.4344 | | 0.369 | 0.7309 | 18600 | 0.5617 | 0.4364 | | 0.369 | 0.7387 | 18800 | 0.5596 | 0.4394 | | 0.3738 | 0.7466 | 19000 | 0.5492 | 0.4292 | | 0.3738 | 0.7545 | 19200 | 0.5478 | 0.4372 | | 0.3738 | 0.7623 | 19400 | 0.5376 | 0.4287 | | 0.368 | 0.7702 | 19600 | 0.5282 | 0.4193 | | 0.368 | 0.7780 | 19800 | 0.5348 | 0.4251 | | 0.3629 | 0.7859 | 20000 | 0.5368 | 0.4313 | | 0.3629 | 0.7937 | 20200 | 0.5551 | 0.4412 | | 0.3629 | 0.8016 | 20400 | 0.5252 | 0.4105 | | 0.3638 | 0.8095 | 20600 | 0.5242 | 0.4117 | | 0.3638 | 0.8173 | 20800 | 0.5233 | 0.4166 | | 0.3512 | 0.8252 | 21000 | 0.5243 | 0.4161 | | 0.3512 | 0.8330 | 21200 | 0.5150 | 0.4123 | | 0.3512 | 0.8409 | 21400 | 0.5089 | 0.4080 | | 0.3536 | 0.8488 | 21600 | 0.5154 | 0.4090 | | 0.3536 | 0.8566 | 21800 | 0.5162 | 0.4092 | | 0.3464 | 0.8645 | 22000 | 0.5098 | 0.4053 | | 0.3464 | 0.8723 | 22200 | 0.5070 | 0.4023 | | 0.3464 | 0.8802 | 22400 | 0.5070 | 0.4071 | | 0.3377 | 0.8881 | 22600 | 0.5028 | 0.3967 | | 0.3377 | 0.8959 | 22800 | 0.5036 | 0.3978 | | 0.3272 | 0.9038 | 23000 | 0.5021 | 0.3954 | | 0.3272 | 0.9116 | 23200 | 0.5033 | 0.3985 | | 0.3272 | 0.9195 | 23400 | 0.4984 | 0.3972 | | 0.319 | 0.9273 | 23600 | 0.4929 | 0.3924 | | 0.319 | 0.9352 | 23800 | 0.4941 | 0.4013 | | 0.3184 | 0.9431 | 24000 | 0.4856 | 0.3874 | | 0.3184 | 0.9509 | 24200 | 0.4892 | 0.3914 | | 0.3184 | 0.9588 | 24400 | 0.4860 | 0.3814 | | 0.3091 | 0.9666 | 24600 | 0.4825 | 0.3834 | | 0.3091 | 0.9745 | 24800 | 0.4784 | 0.3867 | | 0.3154 | 0.9824 | 25000 | 0.4751 | 0.3808 | | 0.3154 | 0.9902 | 25200 | 0.4779 | 0.3849 | | 0.3154 | 0.9981 | 25400 | 0.4773 | 0.3808 | | 0.312 | 1.0059 | 25600 | 0.4777 | 0.3758 | | 0.312 | 1.0138 | 25800 | 0.4752 | 0.3821 | | 0.2651 | 1.0217 | 26000 | 0.4701 | 0.3775 | | 0.2651 | 1.0295 | 26200 | 0.4701 | 0.3761 | | 0.2651 | 1.0374 | 26400 | 0.4718 | 0.3776 | | 0.2627 | 1.0452 | 26600 | 0.4638 | 0.3730 | | 0.2627 | 1.0531 | 26800 | 0.4677 | 0.3720 | | 0.2427 | 1.0609 | 27000 | 0.4643 | 0.3699 | | 0.2427 | 1.0688 | 27200 | 0.4602 | 0.3713 | | 0.2427 | 1.0767 | 27400 | 0.4664 | 0.3703 | | 0.2464 | 1.0845 | 27600 | 0.4609 | 0.3677 | | 0.2464 | 1.0924 | 27800 | 0.4614 | 0.3687 | | 0.2537 | 1.1002 | 28000 | 0.4555 | 0.3655 | | 0.2537 | 1.1081 | 28200 | 0.4560 | 0.3645 | | 0.2537 | 1.1160 | 28400 | 0.4543 | 0.3626 | | 0.2313 | 1.1238 | 28600 | 0.4540 | 0.3631 | | 0.2313 | 1.1317 | 28800 | 0.4536 | 0.3626 | | 0.2451 | 1.1395 | 29000 | 0.4529 | 0.3617 | | 0.2451 | 1.1474 | 29200 | 0.4530 | 0.3598 | | 0.2451 | 1.1553 | 29400 | 0.4515 | 0.3592 | | 0.2445 | 1.1631 | 29600 | 0.4514 | 0.3590 | | 0.2445 | 1.1710 | 29800 | 0.4514 | 0.3589 | | 0.2364 | 1.1788 | 30000 | 0.4511 | 0.3591 | ### Framework versions - Transformers 4.41.2 - Pytorch 2.3.1+cu121 - Datasets 2.20.0 - Tokenizers 0.19.1