wav2vec2-large-xls-r-300m-malay-colab
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.2360
- Wer: 0.5813
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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- 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 |
|---|---|---|---|---|
| 5.2276 | 4.2553 | 400 | 2.8672 | 1.0 |
| 2.7666 | 8.5106 | 800 | 2.0710 | 0.9977 |
| 1.7867 | 12.7660 | 1200 | 1.2444 | 0.8269 |
| 1.0302 | 17.0213 | 1600 | 1.0305 | 0.6500 |
| 0.6215 | 21.2766 | 2000 | 1.1663 | 0.6100 |
| 0.4241 | 25.5319 | 2400 | 1.1573 | 0.5767 |
| 0.3113 | 29.7872 | 2800 | 1.2360 | 0.5813 |
Framework versions
- Transformers 4.49.0.dev0
- Pytorch 2.5.1+cu121
- Datasets 3.2.1.dev0
- Tokenizers 0.21.0
- Downloads last month
- 5
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Model tree for imhakim1/wav2vec2-large-xls-r-300m-malay-colab
Base model
facebook/wav2vec2-xls-r-300m