ntnu-smil/whisper-large-v3-sandi-train-dev-7-merged

whisper-large-v3-sandi-train-dev-7

This model is a fine-tuned version of ntnu-smil/whisper-large-v3-sandi-train-dev-1-merged on the ntnu-smil/sandi2025-ds dataset. It achieves the following results on the evaluation set:

• Loss: 0.9729
• Wer: 65.9380
• Cer: 232.0458
• Decode Runtime: 282.3959
• Wer Runtime: 0.2153
• Cer Runtime: 0.5193

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
1.7464	1.0357	7	1.2461	53.6691	238.2170	279.6409	0.2214	0.5469
1.1723	2.0714	14	1.1018	58.4811	247.6960	282.8760	0.2141	0.5390
1.021	3.1071	21	1.0080	59.1625	246.0458	279.0613	0.2206	0.5396
0.9799	4.1429	28	0.9729	65.9380	232.0458	282.3959	0.2153	0.5193

Framework versions

• PEFT 0.15.1
• Transformers 4.50.3
• Pytorch 2.4.1+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1