<!-- 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_Indonesia_6
This model is a fine-tuned version of facebook/wav2vec2-base-100h on the common_voice dataset. It achieves the following results on the evaluation set:
- Loss: 1.7559
- Wer: 1.0232
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 |
---|---|---|---|---|
3.1807 | 3.23 | 400 | 1.3655 | 1.0052 |
0.5608 | 6.45 | 800 | 1.3604 | 1.0312 |
0.3302 | 9.68 | 1200 | 1.3724 | 1.0355 |
0.2405 | 12.9 | 1600 | 1.4350 | 1.0142 |
0.1883 | 16.13 | 2000 | 1.5079 | 1.0213 |
0.1535 | 19.35 | 2400 | 1.5038 | 1.0251 |
0.1307 | 22.58 | 2800 | 1.7026 | 1.0189 |
0.1104 | 25.81 | 3200 | 1.7072 | 1.0090 |
0.0921 | 29.03 | 3600 | 1.7559 | 1.0232 |
Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu113
- Datasets 1.18.3
- Tokenizers 0.10.3