wav2vec2-base-ks-padpt800

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

  • Loss: 1.5281
  • Accuracy: 0.6142

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: 0.003
  • train_batch_size: 256
  • eval_batch_size: 256
  • seed: 0
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 1024
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 10.0
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.328 1.0 50 1.5281 0.6142
0.9328 2.0 100 1.3054 0.5853
0.8277 3.0 150 1.3858 0.4966
0.7689 4.0 200 1.4112 0.4975
0.7154 5.0 250 1.4042 0.5035
0.706 6.0 300 1.3635 0.5171
0.6878 7.0 350 1.4373 0.4873
0.6868 8.0 400 1.2890 0.5505
0.6705 9.0 450 1.3019 0.5405
0.6579 10.0 500 1.3337 0.5272

Framework versions

  • Transformers 4.22.0.dev0
  • Pytorch 1.11.0+cu115
  • Datasets 2.4.0
  • Tokenizers 0.12.1
Downloads last month
156
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.

Dataset used to train Jungwoo4021/wav2vec2-base-ks-padpt800