KasuleTrevor/cdli-whisper-en-ug-ke-sunbird-encoder-a40

<!-- 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. -->

cdli-whisper-en-ug-ke-sunbird-encoder-a40

This model is a fine-tuned version of Sunbird/asr-whisper-large-v3-salt on the None dataset. It achieves the following results on the evaluation set:

• —Loss: 0.8885
• —Wer: 0.2945
• —Cer: 0.2021
• —test_cer = 0.1224
• —test_loss = 0.6973
• —test_runtime = 0:23:36.28
• —testsamplesper_second = 1.369
• —teststepsper_second = 0.342
• —test_wer = 0.1977

Ugandan English

• —utteranceavgwer = 0.238785
• —utteranceavgcer = 0.144570
• —wer = 0.255675
• —cer = 0.151887

Kenyan English

• —utteranceavgwer = 0.191583
• —utteranceavgcer = 0.114696
• —wer = 0.197774
• —cer = 0.119329

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: 2e-05
• —trainbatchsize: 4
• —evalbatchsize: 4
• —seed: 42
• —gradientaccumulationsteps: 4
• —totaltrainbatch_size: 16
• —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: 150
• —training_steps: 2500

Training results

Framework versions

• —Transformers 4.52.0
• —Pytorch 2.7.1+cu118
• —Datasets 3.6.0
• —Tokenizers 0.21.4