whisper-tiny-indo

This model is a fine-tuned version of openai/whisper-tiny on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 0.3945
• Wer: 32.9085
• Cer: 15.0217

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: 32
• eval_batch_size: 16
• seed: 42
• 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: 500
• training_steps: 5180
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.6158	1.0998	1036	0.4683	36.4109	15.8020
0.3037	3.0994	2072	0.4146	33.6021	14.0364
0.1564	5.0990	3108	0.4005	31.8769	13.0228
0.1035	7.0986	4144	0.3990	31.6948	13.7883
0.0765	9.0983	5180	0.3945	32.9085	15.0217

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.5
• Tokenizers 0.22.2