<!-- 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. -->
w2v2-libri-10min
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.1746
- Wer: 0.5906
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: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 2500
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
5.0324 | 125.0 | 250 | 2.9088 | 1.0 |
0.5241 | 250.0 | 500 | 1.6297 | 0.6030 |
0.0296 | 375.0 | 750 | 1.8666 | 0.6321 |
0.0182 | 500.0 | 1000 | 2.0439 | 0.6473 |
0.0081 | 625.0 | 1250 | 2.0702 | 0.6349 |
0.0079 | 750.0 | 1500 | 2.1576 | 0.6335 |
0.0053 | 875.0 | 1750 | 1.9883 | 0.6376 |
0.0025 | 1000.0 | 2000 | 2.1566 | 0.6072 |
0.002 | 1125.0 | 2250 | 2.1848 | 0.5975 |
0.0015 | 1250.0 | 2500 | 2.1746 | 0.5906 |
Framework versions
- Transformers 4.29.2
- Pytorch 2.0.1+cu118
- Datasets 1.18.3
- Tokenizers 0.13.3