whisper-small-Split-Sentences

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

• Loss: 0.8577
• Cer: 19.6464
• Wer: 35.1503

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: 5e-06
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• 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
• num_epochs: 8
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer	Wer
0.8021	1.0	2706	0.6869	31.9660	54.4888
0.5357	2.0	5412	0.6727	24.5236	41.2723
0.3734	3.0	8118	0.6916	26.2098	42.4365
0.2564	4.0	10824	0.7168	22.5704	36.8897
0.1778	5.0	13530	0.7562	21.5200	36.1022
0.1246	6.0	16236	0.7975	20.6290	36.0816
0.09	7.0	18942	0.8342	20.3819	35.1092
0.0695	8.0	21648	0.8577	19.6464	35.1503

Framework versions

• Transformers 4.56.1
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.22.0