gary109's picture
update model card README.md
b5e33f8
metadata
license: apache-2.0
tags:
  - automatic-speech-recognition
  - gary109/AI_Light_Dance
  - generated_from_trainer
model-index:
  - name: ai-light-dance_stepmania_ft_wav2vec2-large-xlsr-53-v3
    results: []

ai-light-dance_stepmania_ft_wav2vec2-large-xlsr-53-v3

This model is a fine-tuned version of gary109/ai-light-dance_stepmania_ft_wav2vec2-large-xlsr-53-v2 on the GARY109/AI_LIGHT_DANCE - ONSET-STEPMANIA2 dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0734
  • Wer: 0.6928

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: 4e-06
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 16
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 10.0
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.9189 1.0 188 1.0770 0.7002
0.9172 2.0 376 1.0780 0.6955
0.9177 3.0 564 1.0824 0.6916
0.9184 4.0 752 1.0734 0.6928
0.9072 5.0 940 1.0841 0.6897
0.9089 6.0 1128 1.0788 0.6870
0.9174 7.0 1316 1.0761 0.6856
0.9072 8.0 1504 1.0776 0.6850
0.9079 9.0 1692 1.0795 0.6852
0.9016 10.0 1880 1.0817 0.6850

Framework versions

  • Transformers 4.21.0.dev0
  • Pytorch 1.11.0+cu113
  • Datasets 2.3.3.dev0
  • Tokenizers 0.12.1