--- library_name: transformers license: mit base_model: facebook/w2v-bert-2.0 tags: - generated_from_trainer metrics: - wer model-index: - name: w2v-bert-2.0-CV_Fleurs_AMMI_ALFFA-sw-20hrs-v1 results: [] --- # w2v-bert-2.0-CV_Fleurs_AMMI_ALFFA-sw-20hrs-v1 This model is a fine-tuned version of [facebook/w2v-bert-2.0](https://huggingface.co/facebook/w2v-bert-2.0) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.5645 - Wer: 0.2500 - Cer: 0.0899 ## 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: 5e-05 - train_batch_size: 4 - eval_batch_size: 2 - seed: 42 - optimizer: Use 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 - num_epochs: 100 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | Cer | |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:| | 1.7972 | 1.0 | 2970 | 0.6806 | 0.2849 | 0.0958 | | 0.4586 | 2.0 | 5940 | 0.6101 | 0.2595 | 0.0922 | | 0.3832 | 3.0 | 8910 | 0.5412 | 0.2290 | 0.0789 | | 0.3513 | 4.0 | 11880 | 0.4830 | 0.2379 | 0.0865 | | 0.3284 | 5.0 | 14850 | 0.5698 | 0.2259 | 0.0800 | | 0.3268 | 6.0 | 17820 | 0.6145 | 0.2308 | 0.0810 | | 0.3129 | 7.0 | 20790 | 0.5390 | 0.2517 | 0.0883 | | 0.2935 | 8.0 | 23760 | 0.6146 | 0.2366 | 0.0858 | | 0.2829 | 9.0 | 26730 | 0.6222 | 0.2571 | 0.0892 | | 0.2835 | 10.0 | 29700 | 0.6284 | 0.2480 | 0.0907 | | 0.2709 | 11.0 | 32670 | 0.6553 | 0.2542 | 0.0923 | | 0.2468 | 12.0 | 35640 | 0.6046 | 0.2406 | 0.0868 | | 0.2337 | 13.0 | 38610 | 0.6232 | 0.2411 | 0.0880 | | 0.2037 | 14.0 | 41580 | 0.6318 | 0.2290 | 0.0837 | | 0.2021 | 15.0 | 44550 | 0.5645 | 0.2500 | 0.0899 | ### Framework versions - Transformers 4.46.0 - Pytorch 2.1.0+cu118 - Datasets 3.0.2 - Tokenizers 0.20.1