CoolFace
Modelpublic

Hanhpt23/whisper-base-chinesemed-v1

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
0likes5downloads
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. -->

openai/whisper-base

This model is a fine-tuned version of openai/whisper-base on the Hanhpt23/ChineseMed dataset. It achieves the following results on the evaluation set:

  • —Loss: 5.0851
  • —Wer: 123.8122

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: 0.0001
  • —trainbatchsize: 8
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 100
  • —num_epochs: 20

Training results

Training LossEpochStepValidation LossWer
2.99111.022222.9811126.1019
2.44712.044442.9707127.8764
1.85473.066663.2495108.1282
1.25954.088883.5609127.3039
0.91035.0111103.9172114.0813
0.5936.0133324.2574108.8151
0.47387.0155544.4006108.0137
0.37888.0177764.5577136.2335
0.39169.0199984.6187128.9067
0.314810.0222204.7217121.5799
0.341311.0244424.8141122.3812
0.290312.0266644.8305117.5157
0.304413.0288864.8859129.0212
0.264814.0311084.9314111.5054
0.334315.0333304.9714111.4482
0.269316.0355525.0438109.9599
0.267717.0377745.0470108.0710
0.283418.0399965.0293120.2633
0.219819.0422185.0545123.6978
0.224220.0444405.0851123.8122

Framework versions

  • —Transformers 4.41.1
  • —Pytorch 2.3.0
  • —Datasets 2.19.1
  • —Tokenizers 0.19.1