--- language: - fa license: apache-2.0 base_model: makhataei/Whisper-Small-Common-Voice tags: - fa-asr - generated_from_trainer datasets: - mozilla-foundation/common_voice_15_0 metrics: - wer model-index: - name: Whisper Small Persian results: [] --- # Whisper Small Persian This model is a fine-tuned version of [makhataei/Whisper-Small-Common-Voice](https://huggingface.co/makhataei/Whisper-Small-Common-Voice) on the Common Voice 15.0 dataset. It achieves the following results on the evaluation set: - Loss: 0.8695 - Wer: 50.4804 ## 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: 1e-08 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 40 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - training_steps: 10000 ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | |:-------------:|:-----:|:-----:|:---------------:|:-------:| | 0.0002 | 0.14 | 100 | 0.8627 | 50.5962 | | 0.0002 | 0.28 | 200 | 0.8627 | 50.5945 | | 0.0002 | 0.42 | 300 | 0.8628 | 50.5945 | | 0.0001 | 0.56 | 400 | 0.8631 | 50.2720 | | 0.0001 | 0.7 | 500 | 0.8634 | 50.2869 | | 0.0002 | 0.83 | 600 | 0.8637 | 50.2638 | | 0.0002 | 0.97 | 700 | 0.8638 | 50.2704 | | 0.0002 | 1.11 | 800 | 0.8639 | 50.2935 | | 0.0002 | 1.25 | 900 | 0.8640 | 50.2704 | | 0.0002 | 1.39 | 1000 | 0.8641 | 50.2340 | | 0.0002 | 1.53 | 1100 | 0.8644 | 50.2538 | | 0.0002 | 1.67 | 1200 | 0.8645 | 50.2522 | | 0.0002 | 1.81 | 1300 | 0.8646 | 50.2671 | | 0.0002 | 1.95 | 1400 | 0.8648 | 50.2241 | | 0.0002 | 2.09 | 1500 | 0.8650 | 50.2390 | | 0.0001 | 2.23 | 1600 | 0.8653 | 50.2274 | | 0.0001 | 2.36 | 1700 | 0.8653 | 50.2257 | | 0.0002 | 2.5 | 1800 | 0.8653 | 50.2290 | | 0.0002 | 2.64 | 1900 | 0.8653 | 50.2373 | | 0.0002 | 2.78 | 2000 | 0.8653 | 50.2307 | | 0.0001 | 2.92 | 2100 | 0.8655 | 50.2158 | | 0.0002 | 3.06 | 2200 | 0.8656 | 50.2175 | | 0.0002 | 3.2 | 2300 | 0.8658 | 50.2108 | | 0.0002 | 3.34 | 2400 | 0.8659 | 50.2175 | | 0.0001 | 3.48 | 2500 | 0.8660 | 50.2274 | | 0.0001 | 3.62 | 2600 | 0.8661 | 50.2257 | | 0.0002 | 3.76 | 2700 | 0.8662 | 50.2323 | | 0.0002 | 3.89 | 2800 | 0.8663 | 50.2108 | | 0.0002 | 4.03 | 2900 | 0.8665 | 50.1827 | | 0.0002 | 4.17 | 3000 | 0.8666 | 50.2158 | | 0.0001 | 4.31 | 3100 | 0.8668 | 50.2191 | | 0.0002 | 4.45 | 3200 | 0.8668 | 50.2224 | | 0.0002 | 4.59 | 3300 | 0.8669 | 50.1976 | | 0.0002 | 4.73 | 3400 | 0.8668 | 50.1893 | | 0.0002 | 4.87 | 3500 | 0.8668 | 50.1976 | | 0.0001 | 5.01 | 3600 | 0.8670 | 50.1976 | | 0.0002 | 5.15 | 3700 | 0.8671 | 50.1893 | | 0.0001 | 5.29 | 3800 | 0.8672 | 50.1893 | | 0.0002 | 5.42 | 3900 | 0.8673 | 50.1860 | | 0.0001 | 5.56 | 4000 | 0.8674 | 50.1728 | | 0.0002 | 5.7 | 4100 | 0.8674 | 50.1927 | | 0.0002 | 5.84 | 4200 | 0.8675 | 50.2935 | | 0.0001 | 5.98 | 4300 | 0.8676 | 50.1943 | | 0.0002 | 6.12 | 4400 | 0.8676 | 50.2009 | | 0.0002 | 6.26 | 4500 | 0.8677 | 50.2605 | | 0.0002 | 6.4 | 4600 | 0.8677 | 50.2737 | | 0.0002 | 6.54 | 4700 | 0.8679 | 50.2638 | | 0.0001 | 6.68 | 4800 | 0.8681 | 50.2621 | | 0.0002 | 6.82 | 4900 | 0.8681 | 50.2654 | | 0.0001 | 6.95 | 5000 | 0.8681 | 50.2770 | | 0.0001 | 7.09 | 5100 | 0.8682 | 50.2638 | | 0.0002 | 7.23 | 5200 | 0.8682 | 50.2737 | | 0.0002 | 7.37 | 5300 | 0.8682 | 50.2886 | | 0.0002 | 7.51 | 5400 | 0.8683 | 50.2820 | | 0.0002 | 7.65 | 5500 | 0.8683 | 50.3084 | | 0.0001 | 7.79 | 5600 | 0.8684 | 50.2803 | | 0.0001 | 7.93 | 5700 | 0.8685 | 50.2952 | | 0.0001 | 8.07 | 5800 | 0.8686 | 50.2770 | | 0.0001 | 8.21 | 5900 | 0.8687 | 50.2803 | | 0.0001 | 8.34 | 6000 | 0.8688 | 50.2820 | | 0.0001 | 8.48 | 6100 | 0.8689 | 50.3018 | | 0.0002 | 8.62 | 6200 | 0.8689 | 50.2853 | | 0.0001 | 8.76 | 6300 | 0.8689 | 50.2886 | | 0.0002 | 8.9 | 6400 | 0.8689 | 50.2753 | | 0.0001 | 9.04 | 6500 | 0.8689 | 50.4606 | | 0.0001 | 9.18 | 6600 | 0.8690 | 50.4721 | | 0.0001 | 9.32 | 6700 | 0.8690 | 50.4754 | | 0.0002 | 9.46 | 6800 | 0.8690 | 50.4738 | | 0.0002 | 9.6 | 6900 | 0.8691 | 50.4655 | | 0.0001 | 9.74 | 7000 | 0.8692 | 50.4672 | | 0.0002 | 9.87 | 7100 | 0.8692 | 50.4705 | | 0.0002 | 10.01 | 7200 | 0.8692 | 50.4688 | | 0.0001 | 10.15 | 7300 | 0.8692 | 50.4771 | | 0.0001 | 10.29 | 7400 | 0.8692 | 50.4771 | | 0.0001 | 10.43 | 7500 | 0.8692 | 50.4771 | | 0.0001 | 10.57 | 7600 | 0.8692 | 50.4837 | | 0.0001 | 10.71 | 7700 | 0.8693 | 50.4820 | | 0.0001 | 10.85 | 7800 | 0.8693 | 50.4887 | | 0.0001 | 10.99 | 7900 | 0.8693 | 50.4820 | | 0.0001 | 11.13 | 8000 | 0.8694 | 50.4887 | | 0.0001 | 11.27 | 8100 | 0.8694 | 50.4804 | | 0.0002 | 11.4 | 8200 | 0.8694 | 50.4754 | | 0.0001 | 11.54 | 8300 | 0.8694 | 50.4721 | | 0.0002 | 11.68 | 8400 | 0.8694 | 50.4771 | | 0.0001 | 11.82 | 8500 | 0.8694 | 50.4738 | | 0.0002 | 11.96 | 8600 | 0.8695 | 50.4771 | | 0.0002 | 12.1 | 8700 | 0.8695 | 50.4787 | | 0.0001 | 12.24 | 8800 | 0.8695 | 50.4787 | | 0.0001 | 12.38 | 8900 | 0.8695 | 50.4754 | | 0.0001 | 12.52 | 9000 | 0.8695 | 50.4787 | | 0.0001 | 12.66 | 9100 | 0.8695 | 50.4787 | | 0.0001 | 12.8 | 9200 | 0.8695 | 50.4804 | | 0.0001 | 12.93 | 9300 | 0.8695 | 50.4804 | | 0.0001 | 13.07 | 9400 | 0.8695 | 50.4804 | | 0.0002 | 13.21 | 9500 | 0.8695 | 50.4804 | | 0.0002 | 13.35 | 9600 | 0.8695 | 50.4804 | | 0.0001 | 13.49 | 9700 | 0.8695 | 50.4804 | | 0.0002 | 13.63 | 9800 | 0.8695 | 50.4804 | | 0.0002 | 13.77 | 9900 | 0.8695 | 50.4804 | | 0.0001 | 13.91 | 10000 | 0.8695 | 50.4804 | ### Framework versions - Transformers 4.35.2 - Pytorch 2.0.1+cu117 - Datasets 2.15.0 - Tokenizers 0.15.0