<!-- 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. -->
faiq-wav2vec2-large-xlsr-indo-demo-t4-wandb
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice_11_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4045
- Wer: 0.4343
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 |
---|---|---|---|---|
5.0558 | 2.92 | 400 | 2.7962 | 1.0 |
1.4045 | 5.84 | 800 | 0.4595 | 0.6222 |
0.3902 | 8.76 | 1200 | 0.3857 | 0.5292 |
0.2673 | 11.68 | 1600 | 0.4028 | 0.4966 |
0.2135 | 14.6 | 2000 | 0.3953 | 0.4793 |
0.17 | 17.52 | 2400 | 0.4299 | 0.4700 |
0.1527 | 20.44 | 2800 | 0.4106 | 0.4529 |
0.1301 | 23.36 | 3200 | 0.4122 | 0.4488 |
0.1203 | 26.28 | 3600 | 0.4003 | 0.4371 |
0.1078 | 29.2 | 4000 | 0.4045 | 0.4343 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.0+cu118
- Datasets 2.6.1
- Tokenizers 0.13.3