--- tags: - automatic-speech-recognition - gary109/AI_Light_Dance - generated_from_trainer model-index: - name: ai-light-dance_singing3_ft_pretrain2_wav2vec2-large-xlsr-53 results: [] --- # ai-light-dance_singing3_ft_pretrain2_wav2vec2-large-xlsr-53 This model is a fine-tuned version of [gary109/ai-light-dance_pretrain2_wav2vec2-large-xlsr-53](https://huggingface.co/gary109/ai-light-dance_pretrain2_wav2vec2-large-xlsr-53) on the GARY109/AI_LIGHT_DANCE - ONSET-SINGING3 dataset. It achieves the following results on the evaluation set: - Loss: 2.5726 - Wer: 0.9965 ## 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: 8 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 8 - total_train_batch_size: 64 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 100 - num_epochs: 100.0 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | |:-------------:|:-----:|:----:|:---------------:|:------:| | 2.1979 | 1.0 | 72 | 3.4874 | 0.9922 | | 1.9612 | 2.0 | 144 | 4.5105 | 0.9889 | | 1.8052 | 3.0 | 216 | 4.0819 | 0.9787 | | 1.827 | 4.0 | 288 | 4.4058 | 0.9844 | | 1.6249 | 5.0 | 360 | 4.8895 | 0.9872 | | 1.6285 | 6.0 | 432 | 4.7534 | 0.9837 | | 1.6713 | 7.0 | 504 | 4.5588 | 0.9819 | | 1.777 | 8.0 | 576 | 4.7414 | 0.9860 | | 1.6759 | 9.0 | 648 | 4.5222 | 0.9874 | | 1.5533 | 10.0 | 720 | 3.6153 | 0.9824 | | 1.5053 | 11.0 | 792 | 5.6485 | 0.9907 | | 1.6541 | 12.0 | 864 | 5.4657 | 0.9843 | | 1.5891 | 13.0 | 936 | 4.4534 | 0.9877 | | 1.6809 | 14.0 | 1008 | 4.8404 | 0.9891 | | 1.5506 | 15.0 | 1080 | 3.7248 | 0.9873 | | 1.4754 | 16.0 | 1152 | 4.3970 | 0.9864 | | 1.4576 | 17.0 | 1224 | 4.4771 | 0.9855 | | 1.5614 | 18.0 | 1296 | 3.5102 | 0.9861 | | 1.655 | 19.0 | 1368 | 4.8062 | 0.9863 | | 1.455 | 20.0 | 1440 | 5.0790 | 0.9880 | | 1.4701 | 21.0 | 1512 | 4.2320 | 0.9850 | | 1.495 | 22.0 | 1584 | 5.2132 | 0.9889 | | 1.5676 | 23.0 | 1656 | 4.1597 | 0.9806 | | 1.5837 | 24.0 | 1728 | 5.0981 | 0.9867 | | 1.4437 | 25.0 | 1800 | 3.6791 | 0.9813 | | 1.4684 | 26.0 | 1872 | 4.0667 | 0.9869 | | 1.4617 | 27.0 | 1944 | 5.2262 | 0.9902 | | 1.5259 | 28.0 | 2016 | 4.5763 | 0.9876 | | 1.5127 | 29.0 | 2088 | 3.5297 | 0.9831 | | 1.4726 | 30.0 | 2160 | 3.2379 | 0.9776 | | 1.446 | 31.0 | 2232 | 3.9459 | 0.9835 | | 1.4116 | 32.0 | 2304 | 3.5352 | 0.9815 | | 1.462 | 33.0 | 2376 | 3.3658 | 0.9761 | | 1.498 | 34.0 | 2448 | 3.8620 | 0.9795 | | 1.382 | 35.0 | 2520 | 3.4734 | 0.9798 | | 1.3969 | 36.0 | 2592 | 3.7285 | 0.9780 | | 1.3945 | 37.0 | 2664 | 3.7255 | 0.9820 | | 1.4616 | 38.0 | 2736 | 3.8788 | 0.9812 | | 1.4791 | 39.0 | 2808 | 3.1891 | 0.9781 | | 1.3875 | 40.0 | 2880 | 3.6854 | 0.9801 | | 1.3914 | 41.0 | 2952 | 3.4011 | 0.9742 | | 1.3744 | 42.0 | 3024 | 3.5792 | 0.9808 | | 1.4085 | 43.0 | 3096 | 2.9806 | 0.9743 | | 1.407 | 44.0 | 3168 | 3.3188 | 0.9821 | | 1.3678 | 45.0 | 3240 | 3.6300 | 0.9817 | | 1.3762 | 46.0 | 3312 | 3.0986 | 0.9774 | | 1.3507 | 47.0 | 3384 | 2.8956 | 0.9733 | | 1.3646 | 48.0 | 3456 | 2.8360 | 0.9731 | | 1.4027 | 49.0 | 3528 | 3.5181 | 0.9744 | | 1.3383 | 50.0 | 3600 | 3.2710 | 0.9769 | | 1.3517 | 51.0 | 3672 | 2.9974 | 0.9756 | | 1.3551 | 52.0 | 3744 | 3.3180 | 0.9751 | | 1.3493 | 53.0 | 3816 | 2.6087 | 0.9994 | | 1.3977 | 54.0 | 3888 | 2.6875 | 0.9778 | | 1.3254 | 55.0 | 3960 | 3.3663 | 0.9787 | | 1.303 | 56.0 | 4032 | 2.6154 | 0.9692 | | 1.2612 | 57.0 | 4104 | 3.2773 | 0.9761 | | 1.3161 | 58.0 | 4176 | 2.9020 | 0.9717 | | 1.3026 | 59.0 | 4248 | 3.1517 | 0.9709 | | 1.262 | 60.0 | 4320 | 3.1299 | 0.9675 | | 1.2885 | 61.0 | 4392 | 3.1023 | 0.9724 | | 1.2538 | 62.0 | 4464 | 2.7857 | 0.9877 | | 1.2536 | 63.0 | 4536 | 3.0150 | 0.9704 | | 1.3339 | 64.0 | 4608 | 2.9305 | 0.9722 | | 1.2392 | 65.0 | 4680 | 2.7695 | 0.9749 | | 1.2572 | 66.0 | 4752 | 2.8677 | 0.9689 | | 1.2538 | 67.0 | 4824 | 2.8597 | 0.9805 | | 1.2355 | 68.0 | 4896 | 2.8300 | 0.9788 | | 1.2682 | 69.0 | 4968 | 2.7011 | 0.9742 | | 1.2252 | 70.0 | 5040 | 2.8322 | 0.9722 | | 1.2085 | 71.0 | 5112 | 2.7401 | 0.9711 | | 1.2412 | 72.0 | 5184 | 2.6227 | 0.9763 | | 1.2148 | 73.0 | 5256 | 2.7146 | 0.9690 | | 1.2411 | 74.0 | 5328 | 2.9963 | 0.9663 | | 1.2012 | 75.0 | 5400 | 2.7940 | 0.9655 | | 1.1969 | 76.0 | 5472 | 2.6212 | 0.9790 | | 1.2004 | 77.0 | 5544 | 2.8644 | 0.9711 | | 1.2247 | 78.0 | 5616 | 2.6127 | 0.9621 | | 1.2791 | 79.0 | 5688 | 2.6538 | 0.9719 | | 1.1867 | 80.0 | 5760 | 2.7779 | 0.9587 | | 1.1992 | 81.0 | 5832 | 2.8297 | 0.9643 | | 1.1833 | 82.0 | 5904 | 2.8508 | 0.9723 | | 1.2092 | 83.0 | 5976 | 2.8261 | 0.9690 | | 1.2215 | 84.0 | 6048 | 2.5995 | 0.9717 | | 1.1966 | 85.0 | 6120 | 2.7659 | 0.9677 | | 1.1774 | 86.0 | 6192 | 2.6289 | 0.9703 | | 1.2002 | 87.0 | 6264 | 2.7197 | 0.9775 | | 1.2035 | 88.0 | 6336 | 2.6154 | 0.9778 | | 1.2219 | 89.0 | 6408 | 2.6426 | 0.9761 | | 1.1724 | 90.0 | 6480 | 2.5996 | 0.9801 | | 1.1832 | 91.0 | 6552 | 2.7315 | 0.9743 | | 1.1759 | 92.0 | 6624 | 2.5726 | 0.9965 | | 1.1638 | 93.0 | 6696 | 2.6570 | 0.9865 | | 1.1872 | 94.0 | 6768 | 2.6414 | 0.9948 | | 1.144 | 95.0 | 6840 | 2.6264 | 0.9863 | | 1.1636 | 96.0 | 6912 | 2.5820 | 0.9918 | | 1.1714 | 97.0 | 6984 | 2.5970 | 0.9913 | | 1.1943 | 98.0 | 7056 | 2.6308 | 0.9895 | | 1.185 | 99.0 | 7128 | 2.6379 | 0.9898 | | 1.1569 | 100.0 | 7200 | 2.6331 | 0.9906 | ### Framework versions - Transformers 4.21.0.dev0 - Pytorch 1.9.1+cu102 - Datasets 2.3.3.dev0 - Tokenizers 0.12.1