<!-- 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. -->
pseudolabeling-step2-F01-Pass-2
This model is a fine-tuned version of monideep2255/XLRS-torgo on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.2704
- Wer: 1.1942
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: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.9509 | 2.94 | 400 | 1.1949 | 1.2734 |
| 0.5686 | 5.88 | 800 | 1.0452 | 1.2266 |
| 0.4179 | 8.82 | 1200 | 1.1876 | 1.2032 |
| 0.3137 | 11.76 | 1600 | 1.2691 | 1.2572 |
| 0.2329 | 14.7 | 2000 | 1.2944 | 1.2104 |
| 0.1851 | 17.64 | 2400 | 1.4389 | 1.2626 |
| 0.1427 | 20.59 | 2800 | 1.3325 | 1.2608 |
| 0.1101 | 23.53 | 3200 | 1.4132 | 1.2176 |
| 0.0805 | 26.47 | 3600 | 1.3443 | 1.2482 |
| 0.0645 | 29.41 | 4000 | 1.2704 | 1.1942 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 1.18.3
- Tokenizers 0.13.2
