CoolFace
Modelpublic

gary109/ai-light-dance_stepmania_ft_wav2vec2-large-xlsr-53-v1

sourceHugging Faceapache-2.0updated 4y agoView on Hugging Face
0likes7downloads
Model Card

<!-- 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. -->

ai-light-dancestepmaniaft_wav2vec2-large-xlsr-53-v1

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

  • —Loss: 1.0763
  • —Wer: 0.7344

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-05
  • —trainbatchsize: 10
  • —evalbatchsize: 10
  • —seed: 42
  • —gradientaccumulationsteps: 16
  • —totaltrainbatch_size: 160
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 100
  • —num_epochs: 10.0
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossWer
1.16321.01501.20070.9875
1.16152.03001.19120.9875
1.14873.04501.19420.9875
1.12074.06001.17530.9875
1.06385.07501.13450.8214
1.01746.09001.15410.7665
0.99467.010501.07990.7716
0.96948.012001.08480.7418
0.95669.013501.07630.7344
0.946610.015001.07910.7240

Framework versions

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