--- 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: 0.4908 - Wer: 0.3964 ## 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: 20000 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | |:-------------:|:------:|:-----:|:---------------:|:------:| | No log | 0.0079 | 200 | 3.1856 | 1.0 | | No log | 0.0157 | 400 | 2.6492 | 1.0 | | 4.6997 | 0.0236 | 600 | 1.3869 | 0.8722 | | 4.6997 | 0.0314 | 800 | 1.2302 | 0.8308 | | 1.0569 | 0.0393 | 1000 | 1.1380 | 0.7958 | | 1.0569 | 0.0472 | 1200 | 1.0668 | 0.7698 | | 1.0569 | 0.0550 | 1400 | 1.0208 | 0.7310 | | 0.8131 | 0.0629 | 1600 | 0.9702 | 0.7151 | | 0.8131 | 0.0707 | 1800 | 0.9408 | 0.6882 | | 0.7194 | 0.0786 | 2000 | 0.9250 | 0.6804 | | 0.7194 | 0.0864 | 2200 | 0.9052 | 0.6726 | | 0.7194 | 0.0943 | 2400 | 0.8986 | 0.6573 | | 0.6688 | 0.1022 | 2600 | 0.8815 | 0.6473 | | 0.6688 | 0.1100 | 2800 | 0.8588 | 0.6445 | | 0.645 | 0.1179 | 3000 | 0.8758 | 0.6487 | | 0.645 | 0.1257 | 3200 | 0.8725 | 0.6691 | | 0.645 | 0.1336 | 3400 | 0.8296 | 0.6298 | | 0.6077 | 0.1415 | 3600 | 0.8356 | 0.6552 | | 0.6077 | 0.1493 | 3800 | 0.8263 | 0.6229 | | 0.5983 | 0.1572 | 4000 | 0.8711 | 0.6885 | | 0.5983 | 0.1650 | 4200 | 0.7837 | 0.5918 | | 0.5983 | 0.1729 | 4400 | 0.8097 | 0.6598 | | 0.5788 | 0.1808 | 4600 | 0.7777 | 0.5869 | | 0.5788 | 0.1886 | 4800 | 0.7913 | 0.5896 | | 0.5501 | 0.1965 | 5000 | 0.7924 | 0.5900 | | 0.5501 | 0.2043 | 5200 | 0.7603 | 0.5737 | | 0.5501 | 0.2122 | 5400 | 0.7750 | 0.5932 | | 0.5694 | 0.2200 | 5600 | 0.7517 | 0.5711 | | 0.5694 | 0.2279 | 5800 | 0.7651 | 0.5698 | | 0.5424 | 0.2358 | 6000 | 0.7548 | 0.5820 | | 0.5424 | 0.2436 | 6200 | 0.7305 | 0.5681 | | 0.5424 | 0.2515 | 6400 | 0.7314 | 0.5589 | | 0.521 | 0.2593 | 6600 | 0.7228 | 0.5654 | | 0.521 | 0.2672 | 6800 | 0.7350 | 0.5633 | | 0.5119 | 0.2751 | 7000 | 0.7079 | 0.5347 | | 0.5119 | 0.2829 | 7200 | 0.7105 | 0.5601 | | 0.5119 | 0.2908 | 7400 | 0.6876 | 0.5378 | | 0.5007 | 0.2986 | 7600 | 0.6835 | 0.5303 | | 0.5007 | 0.3065 | 7800 | 0.7132 | 0.5351 | | 0.4934 | 0.3144 | 8000 | 0.6972 | 0.5242 | | 0.4934 | 0.3222 | 8200 | 0.6800 | 0.5227 | | 0.4934 | 0.3301 | 8400 | 0.6916 | 0.5365 | | 0.4762 | 0.3379 | 8600 | 0.6802 | 0.5255 | | 0.4762 | 0.3458 | 8800 | 0.6978 | 0.5337 | | 0.4774 | 0.3536 | 9000 | 0.6567 | 0.5211 | | 0.4774 | 0.3615 | 9200 | 0.6479 | 0.5152 | | 0.4774 | 0.3694 | 9400 | 0.6551 | 0.5147 | | 0.4632 | 0.3772 | 9600 | 0.6358 | 0.4955 | | 0.4632 | 0.3851 | 9800 | 0.6466 | 0.5109 | | 0.4483 | 0.3929 | 10000 | 0.6306 | 0.5044 | | 0.4483 | 0.4008 | 10200 | 0.6360 | 0.5004 | | 0.4483 | 0.4087 | 10400 | 0.6302 | 0.4914 | | 0.4454 | 0.4165 | 10600 | 0.6163 | 0.4851 | | 0.4454 | 0.4244 | 10800 | 0.6221 | 0.4911 | | 0.4302 | 0.4322 | 11000 | 0.6396 | 0.5001 | | 0.4302 | 0.4401 | 11200 | 0.6212 | 0.4841 | | 0.4302 | 0.4480 | 11400 | 0.6268 | 0.4938 | | 0.4261 | 0.4558 | 11600 | 0.6098 | 0.4820 | | 0.4261 | 0.4637 | 11800 | 0.6009 | 0.4689 | | 0.4026 | 0.4715 | 12000 | 0.6091 | 0.4810 | | 0.4026 | 0.4794 | 12200 | 0.6019 | 0.4806 | | 0.4026 | 0.4872 | 12400 | 0.5947 | 0.4671 | | 0.4027 | 0.4951 | 12600 | 0.5994 | 0.4709 | | 0.4027 | 0.5030 | 12800 | 0.5982 | 0.4761 | | 0.3978 | 0.5108 | 13000 | 0.5890 | 0.4632 | | 0.3978 | 0.5187 | 13200 | 0.5871 | 0.4567 | | 0.3978 | 0.5265 | 13400 | 0.5873 | 0.4635 | | 0.3875 | 0.5344 | 13600 | 0.5772 | 0.4539 | | 0.3875 | 0.5423 | 13800 | 0.5604 | 0.4419 | | 0.404 | 0.5501 | 14000 | 0.5689 | 0.4454 | | 0.404 | 0.5580 | 14200 | 0.5595 | 0.4433 | | 0.404 | 0.5658 | 14400 | 0.5575 | 0.4406 | | 0.3878 | 0.5737 | 14600 | 0.5522 | 0.4353 | | 0.3878 | 0.5816 | 14800 | 0.5522 | 0.4352 | | 0.3622 | 0.5894 | 15000 | 0.5570 | 0.4401 | | 0.3622 | 0.5973 | 15200 | 0.5467 | 0.4280 | | 0.3622 | 0.6051 | 15400 | 0.5511 | 0.4340 | | 0.3545 | 0.6130 | 15600 | 0.5437 | 0.4245 | | 0.3545 | 0.6208 | 15800 | 0.5489 | 0.4296 | | 0.3486 | 0.6287 | 16000 | 0.5420 | 0.4278 | | 0.3486 | 0.6366 | 16200 | 0.5352 | 0.4213 | | 0.3486 | 0.6444 | 16400 | 0.5377 | 0.4259 | | 0.3374 | 0.6523 | 16600 | 0.5336 | 0.4305 | | 0.3374 | 0.6601 | 16800 | 0.5294 | 0.4188 | | 0.3389 | 0.6680 | 17000 | 0.5253 | 0.4169 | | 0.3389 | 0.6759 | 17200 | 0.5194 | 0.4144 | | 0.3389 | 0.6837 | 17400 | 0.5232 | 0.4171 | | 0.3258 | 0.6916 | 17600 | 0.5179 | 0.4165 | | 0.3258 | 0.6994 | 17800 | 0.5132 | 0.4104 | | 0.327 | 0.7073 | 18000 | 0.5096 | 0.4044 | | 0.327 | 0.7152 | 18200 | 0.5041 | 0.4034 | | 0.327 | 0.7230 | 18400 | 0.5013 | 0.3981 | | 0.316 | 0.7309 | 18600 | 0.5074 | 0.4065 | | 0.316 | 0.7387 | 18800 | 0.5014 | 0.4055 | | 0.3162 | 0.7466 | 19000 | 0.4959 | 0.3998 | | 0.3162 | 0.7545 | 19200 | 0.4930 | 0.3982 | | 0.3162 | 0.7623 | 19400 | 0.4925 | 0.3982 | | 0.3145 | 0.7702 | 19600 | 0.4922 | 0.3970 | | 0.3145 | 0.7780 | 19800 | 0.4908 | 0.3969 | | 0.3095 | 0.7859 | 20000 | 0.4908 | 0.3964 | ### Framework versions - Transformers 4.41.2 - Pytorch 2.3.1+cu121 - Datasets 2.20.0 - Tokenizers 0.19.1