CoolFace
Modelpublic

arun100/whisper-small-derived-hi-2

sourceHugging Faceapache-2.0updated 3y agoView on Hugging Face
0likes16downloads
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. -->

Whisper Base Bengali

This model is a fine-tuned version of arun100/whisper-small-derived-hi-1 on the mozilla-foundation/commonvoice16_0 hi dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.1458
  • —Wer: 9.2881

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: 5e-07
  • —trainbatchsize: 16
  • —evalbatchsize: 16
  • —seed: 42
  • —gradientaccumulationsteps: 4
  • —totaltrainbatch_size: 64
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 20
  • —training_steps: 1000
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossWer
0.13681.021000.16659.6951
0.11962.052000.15869.6011
0.1073.073000.15389.4008
0.10515.024000.15049.3235
0.09886.055000.14869.4467
0.09397.076000.14749.4425
0.09019.027000.14649.3006
0.085910.048000.14599.4362
0.085911.079000.14589.2881
0.083913.0210000.14569.2901

Framework versions

  • —Transformers 4.38.0.dev0
  • —Pytorch 2.1.2+cu121
  • —Datasets 2.16.2.dev0
  • —Tokenizers 0.15.0