CoolFace
Modelpublic

tz579/example_asr_wav2vec2

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
0likes8downloads
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. -->

facebook/wav2vec2-large-lv60

This model is a fine-tuned version of facebook/wav2vec2-large-lv60 on the EDINBURGHCSTR/AMI - IHM dataset. It achieves the following results on the evaluation set:

  • —Loss: 1.2723
  • —Wer: 0.9542

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: 0.0003
  • —trainbatchsize: 16
  • —evalbatchsize: 16
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 500
  • —num_epochs: 2.0
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossWer
1.09190.156510001.01690.7064
1.47680.313120000.71560.4356
0.97280.469630000.64620.4030
0.54180.626240000.61710.3707
0.84920.782750000.57580.3695
1.48260.939360000.58010.3545
0.32741.095870000.52440.3375
0.20891.252380000.50470.3239
0.29161.408990000.49010.3171
0.16171.5654100000.50700.3151
0.38151.7220110000.49480.3180
1.01711.8785120000.94650.8379

Framework versions

  • —Transformers 4.42.0.dev0
  • —Pytorch 2.3.0a0+gitcd033a1
  • —Datasets 2.19.1
  • —Tokenizers 0.19.1