IoanaLiviaPopescu/real-data-whisper-small-without-normalize-eval

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

• Loss: 0.4132
• Wer: 26.8763

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_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0	0	0.6024	35.6728
0.501	1.0	13	0.4965	28.2987
0.3039	2.0	26	0.4347	28.6356
0.1973	3.0	39	0.4132	26.8763
0.1431	4.0	52	0.4076	26.9699
0.1181	5.0	65	0.4097	27.1196

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1