CoolFace
Modelpublic

alpcansoydas/whisper-large-v2-tr-ft-24-03-26-encoder-only-25ksamples-simulated

sourceHugging Faceapache-2.0updated 6mo 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. -->

Whisper for Turkish Call Centers

This model is a fine-tuned version of openai/whisper-large-v2 on the Custom turkish call center simulated data dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.2663

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-06
  • —trainbatchsize: 16
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 500
  • —training_steps: 4500
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation Loss
0.31060.17772500.3197
0.26590.35545000.2988
0.27730.53307500.2874
0.29770.710710000.2810
0.27560.888412500.2781
0.25311.066115000.2753
0.24901.243817500.2741
0.24451.421520000.2706
0.24301.599122500.2694
0.31671.776825000.2686
0.25141.954527500.2665
0.24452.132230000.2702
0.23652.309932500.2675
0.18002.487635000.2667
0.24212.665237500.2660
0.22822.842940000.2656
0.23063.020642500.2653
0.21743.198345000.2663

Framework versions

  • —Transformers 5.3.0
  • —Pytorch 2.10.0+cu128
  • —Datasets 4.8.4
  • —Tokenizers 0.22.2