CoolFace
Modelpublic

gary109/ai-light-dance_stepmania_ft_wav2vec2-large-xlsr-53-v4

sourceHugging Faceapache-2.0updated 4y agoView on Hugging Face
0likes18downloads
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-v4

This model is a fine-tuned version of gary109/ai-light-dance_stepmania_ft_wav2vec2-large-xlsr-53-v3 on the GARY109/AILIGHTDANCE - ONSET-STEPMANIA2 dataset. It achieves the following results on the evaluation set:

  • —Loss: 1.0298
  • —Wer: 0.6642

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: 8
  • —evalbatchsize: 8
  • —seed: 42
  • —gradientaccumulationsteps: 16
  • —totaltrainbatch_size: 128
  • —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
0.92181.01881.07180.6958
0.91942.03761.03540.6937
0.90773.05641.03650.6730
0.89564.07521.04970.6727
0.8775.09401.02990.6694
0.87366.011281.02980.6642
0.87697.013161.03480.6584
0.85718.015041.06890.6602
0.85739.016921.05590.6549
0.845810.018801.07060.6588

Framework versions

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