Edit model card

ft_wav2vec2_base_thirty

This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6535
  • Wer: 32.0197
  • Cer: 14.4551

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: 0.0001
  • train_batch_size: 32
  • eval_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 2000
  • training_steps: 2700

Training results

Training Loss Epoch Step Validation Loss Wer Cer
4.2536 1.75 500 3.0201 99.9919 99.8547
2.7796 3.51 1000 1.8276 90.7584 46.9070
0.9842 5.26 1500 0.7994 43.5201 19.4787
0.595 7.02 2000 0.7414 37.7699 17.0733
0.4215 8.77 2500 0.6535 32.0197 14.4551

Framework versions

  • Transformers 4.39.3
  • Pytorch 1.12.1+cu116
  • Datasets 2.18.0
  • Tokenizers 0.15.2
Downloads last month
0
Safetensors
Model size
94.4M params
Tensor type
F32
·

Finetuned from