ntnu-smil/whisper-large-v3-sandi-train-dev-1-ex-transcript-merged

whisper-large-v3-sandi-train-dev-1-ex-transcript

This model is a fine-tuned version of openai/whisper-large-v3 on the ntnu-smil/sandi2025-ds dataset. It achieves the following results on the evaluation set:

• Loss: 1.1780
• Wer: 82.0272
• Cer: 240.5125
• Decode Runtime: 292.3134
• Wer Runtime: 0.2232
• Cer Runtime: 0.5411

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: 7e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 1024
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.98) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• training_steps: 28

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer	Decode Runtime	Wer Runtime	Cer Runtime
2.2505	1.0357	7	1.6079	52.4500	235.9007	270.0744	0.2180	0.5443
1.4494	2.0714	14	1.3797	70.5252	246.6679	293.3539	0.2341	0.5711
1.2504	3.1071	21	1.2363	81.4127	248.2291	273.7435	0.2221	0.5429
1.177	4.1429	28	1.1780	82.0272	240.5125	292.3134	0.2232	0.5411

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.4.1+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1