metadata
license: mit
base_model: TheAIchemist13/hindi_beekeeping_wav2vec2
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: hindi_wav2vec2_optimized
results: []
hindi_wav2vec2_optimized
This model is a fine-tuned version of TheAIchemist13/hindi_beekeeping_wav2vec2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6848
- Wer: 0.4059
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: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.312 | 0.11 | 25 | 0.7359 | 0.4045 |
| 0.2789 | 0.22 | 50 | 0.7344 | 0.4129 |
| 0.3066 | 0.33 | 75 | 0.7687 | 0.4185 |
| 0.446 | 0.44 | 100 | 0.9725 | 0.4731 |
| 0.4249 | 0.54 | 125 | 1.2325 | 0.5073 |
| 1.2771 | 0.65 | 150 | 1.1657 | 0.5297 |
| 0.4671 | 0.76 | 175 | 0.9309 | 0.4759 |
| 0.759 | 0.87 | 200 | 1.0748 | 0.5038 |
| 0.4854 | 0.98 | 225 | 0.9735 | 0.4990 |
| 0.4588 | 1.09 | 250 | 0.8566 | 0.4717 |
| 1.429 | 1.2 | 275 | 0.8819 | 0.5073 |
| 0.495 | 1.31 | 300 | 0.8920 | 0.4710 |
| 0.3768 | 1.42 | 325 | 0.9280 | 0.4885 |
| 0.4792 | 1.53 | 350 | 0.9338 | 0.4689 |
| 0.4219 | 1.63 | 375 | 1.0726 | 0.5017 |
| 0.4818 | 1.74 | 400 | 0.9469 | 0.4913 |
| 0.4148 | 1.85 | 425 | 0.9698 | 0.5108 |
| 0.5901 | 1.96 | 450 | 0.9540 | 0.5129 |
| 0.4976 | 2.07 | 475 | 1.0502 | 0.5395 |
| 0.5696 | 2.18 | 500 | 0.8864 | 0.4675 |
| 0.4342 | 2.29 | 525 | 0.8964 | 0.4661 |
| 0.3418 | 2.4 | 550 | 0.9535 | 0.4941 |
| 0.5812 | 2.51 | 575 | 0.8451 | 0.4514 |
| 0.3522 | 2.61 | 600 | 0.9054 | 0.4864 |
| 0.4135 | 2.72 | 625 | 0.8788 | 0.4493 |
| 0.3061 | 2.83 | 650 | 0.8583 | 0.4570 |
| 0.4118 | 2.94 | 675 | 0.8105 | 0.4605 |
| 0.3642 | 3.05 | 700 | 0.8026 | 0.4409 |
| 0.2803 | 3.16 | 725 | 0.8131 | 0.4297 |
| 0.3533 | 3.27 | 750 | 0.7614 | 0.4346 |
| 0.2616 | 3.38 | 775 | 0.8177 | 0.4535 |
| 0.3203 | 3.49 | 800 | 0.7776 | 0.4409 |
| 0.2572 | 3.59 | 825 | 0.7438 | 0.4206 |
| 0.312 | 3.7 | 850 | 0.7454 | 0.4143 |
| 0.2338 | 3.81 | 875 | 0.7815 | 0.4192 |
| 0.2921 | 3.92 | 900 | 0.7109 | 0.4122 |
| 0.2568 | 4.03 | 925 | 0.7430 | 0.4157 |
| 0.1986 | 4.14 | 950 | 0.7301 | 0.4101 |
| 0.2569 | 4.25 | 975 | 0.7449 | 0.4150 |
| 0.1871 | 4.36 | 1000 | 0.7674 | 0.4108 |
| 0.2449 | 4.47 | 1025 | 0.7218 | 0.4080 |
| 0.1846 | 4.58 | 1050 | 0.7219 | 0.3933 |
| 0.2327 | 4.68 | 1075 | 0.7094 | 0.4080 |
| 0.1921 | 4.79 | 1100 | 0.6858 | 0.4031 |
| 0.2246 | 4.9 | 1125 | 0.6848 | 0.4059 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1