<!-- 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-tiengviet_mydata
This model is a fine-tuned version of nguyenvulebinh/wav2vec2-base-vietnamese-250h on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.8207
- Wer: 0.8746
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: 300
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
2.9168 | 23.53 | 400 | 1.5943 | 1.0044 |
0.9836 | 47.06 | 800 | 1.3882 | 0.9252 |
0.628 | 70.59 | 1200 | 1.4862 | 0.9432 |
0.46 | 94.12 | 1600 | 1.5574 | 0.9441 |
0.3498 | 117.65 | 2000 | 1.7666 | 0.9006 |
0.2652 | 141.18 | 2400 | 2.3260 | 0.9067 |
0.205 | 164.71 | 2800 | 2.3611 | 0.8711 |
0.1826 | 188.24 | 3200 | 2.3841 | 0.8891 |
0.1413 | 211.76 | 3600 | 2.3994 | 0.9050 |
0.1026 | 235.29 | 4000 | 2.4840 | 0.8702 |
0.0869 | 258.82 | 4400 | 2.6500 | 0.8781 |
0.082 | 282.35 | 4800 | 2.8207 | 0.8746 |
Framework versions
- Transformers 4.27.3
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2