<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
wav2vec2-common_voice-tr-demo
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the COMMON_VOICE - TR dataset. It achieves the following results on the evaluation set:
- Loss: 0.3997
- Wer: 0.3538
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: 500
- num_epochs: 15.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 0.92 | 100 | 3.6061 | 1.0 |
| No log | 1.83 | 200 | 3.0203 | 0.9999 |
| No log | 2.75 | 300 | 0.9479 | 0.7916 |
| No log | 3.67 | 400 | 0.6024 | 0.6285 |
| 3.1561 | 4.59 | 500 | 0.5112 | 0.5369 |
| 3.1561 | 5.5 | 600 | 0.4581 | 0.4900 |
| 3.1561 | 6.42 | 700 | 0.4321 | 0.4633 |
| 3.1561 | 7.34 | 800 | 0.4252 | 0.4400 |
| 3.1561 | 8.26 | 900 | 0.4204 | 0.4229 |
| 0.2247 | 9.17 | 1000 | 0.3948 | 0.3971 |
| 0.2247 | 10.09 | 1100 | 0.3997 | 0.3963 |
| 0.2247 | 11.01 | 1200 | 0.4157 | 0.3894 |
| 0.2247 | 11.93 | 1300 | 0.4142 | 0.3855 |
| 0.2247 | 12.84 | 1400 | 0.4108 | 0.3638 |
| 0.1022 | 13.76 | 1500 | 0.3929 | 0.3618 |
| 0.1022 | 14.68 | 1600 | 0.4004 | 0.3544 |
Framework versions
- Transformers 4.28.1
- Pytorch 1.12.1+cu102
- Datasets 2.12.0
- Tokenizers 0.13.3
