metadata
tags:
- generated_from_trainer
datasets:
- common_voice_17_0
model-index:
- name: wav2vec2-large-xls-r-300m-ipa
results: []
wav2vec2-large-xls-r-300m-ipa
This model was trained from scratch on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.7900
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: 34
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 68
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 7.2967 | 2.2222 | 50 | 7.3217 |
| 3.9212 | 4.4444 | 100 | 4.1064 |
| 3.2366 | 6.6667 | 150 | 3.3641 |
| 3.1466 | 8.8889 | 200 | 3.3592 |
| 3.0562 | 11.1111 | 250 | 3.1385 |
| 0.6813 | 13.3333 | 300 | 0.7900 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1