metadata

library_name: transformers
language:
  - hr
license: apache-2.0
base_model: GoranS/whisper-large-v3-turbo-hr-parla
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: whisper-large-v3-turbo.hr
    results: []

whisper-large-v3-turbo.hr

This model is a fine-tuned version of GoranS/whisper-large-v3-turbo-hr-parla on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1473
Wer: 0.1155

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: 6.25e-06
train_batch_size: 64
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: 500
num_epochs: 2
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1629	0.5450	1000	0.1638	0.1151
0.1133	1.0899	2000	0.1513	0.1168
0.0979	1.6349	3000	0.1473	0.1155

Framework versions

Transformers 4.46.3
Pytorch 2.5.0+cu121
Datasets 3.1.0
Tokenizers 0.20.3