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

Loss: 0.5964
Wer: 0.1869

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: 16
eval_batch_size: 16
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: 20
training_steps: 400
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.3484	2.2727	50	1.0828	0.1986
0.7309	4.5455	100	0.9440	0.1752
0.5176	6.8182	150	0.9446	0.1869
0.2552	9.0909	200	0.5440	0.1923
0.0315	11.3636	250	0.5883	0.1959
0.0069	13.6364	300	0.5849	0.1842
0.0041	15.9091	350	0.5954	0.1851
0.0033	18.1818	400	0.5964	0.1869

Framework versions

Transformers 4.52.0.dev0
Pytorch 2.6.0+cu124
Datasets 3.5.0
Tokenizers 0.21.1

archieto0412/whisper-atc-finetuned

Fine-tuned Whisper for AM

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

archieto0412/
whisper-atc-finetuned