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
• test_samples_per_second = 1.369
• test_steps_per_second = 0.342
• test_wer = 0.1977

Ugandan English

• utterance_avg_wer = 0.238785
• utterance_avg_cer = 0.144570
• wer = 0.255675
• cer = 0.151887

Kenyan English

• utterance_avg_wer = 0.191583
• utterance_avg_cer = 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
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_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