This model is a fine-tuned version of openai/whisper-medium.en on the xbilek25/train_hall_absorb_0.1_speed15 dataset. It achieves the following results on the evaluation set:

Loss: 0.8573
Wer: 26.1176

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: 48
eval_batch_size: 32
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: 375
training_steps: 2250
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0	0	2.1191	32.3637
0.631	0.1667	375	0.8755	35.5175
0.4428	0.3333	750	0.8454	27.8628
0.245	1.1667	1125	0.8499	26.9137
0.1727	1.3333	1500	0.8380	27.4954
0.0979	2.1667	1875	0.8513	26.8524
0.078	2.3333	2250	0.8573	26.1176

Framework versions

Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.1
Tokenizers 0.21.1

xbilek25/whisper-medium-en-cv-9.0

whisper-medium-en-cv-9.0

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

xbilek25/
whisper-medium-en-cv-9.0