Cafet/whisper-small-mongolian

<!-- 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-small-mongolian

• —eval_loss: 0.0255
• —eval_wer: 3.9864
• —eval_runtime: 1588.8988
• —evalsamplesper_second: 1.17
• —evalstepsper_second: 0.147
• —epoch: 2.3112
• —step: 1500

Training hyperparameters

The following hyperparameters were used during training:

• —learning_rate: 1e-05
• —trainbatchsize: 16
• —evalbatchsize: 8
• —seed: 42
• —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• —lrschedulertype: linear
• —lrschedulerwarmup_steps: 1000
• —training_steps: 3000
• —mixedprecisiontraining: Native AMP

Framework versions

• —Transformers 4.40.1
• —Pytorch 2.2.1+cu121
• —Datasets 2.19.0
• —Tokenizers 0.19.1