rbcurzon/whisper-small-ph

whisper-medium-test

This model is a fine-tuned version of openai/whisper-small on the rbcurzon/ph_dialect_asr all dataset. It achieves the following results on the evaluation set:

• Loss: 0.4044
• Wer: 0.1675

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: 1e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• total_eval_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH 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: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2023	2.9607	1000	0.3294	0.1874
0.0309	5.9192	2000	0.3414	0.1717
0.0037	8.8777	3000	0.3782	0.1708
0.0018	11.8362	4000	0.3964	0.1680
0.0013	14.7947	5000	0.4044	0.1675

Framework versions

• Transformers 4.56.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2