metadata

library_name: transformers
language:
  - pa
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: Punjabi Whisper large-v3  - Swayangjit
    results: []

Punjabi Whisper large-v3 - Swayangjit

This model is a fine-tuned version of openai/whisper-large-v3 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.3908
Wer: 71.4286

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: 0.0001
train_batch_size: 16
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: 50
training_steps: 50
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.4502	0.0133	10	0.6460	91.9414
0.7124	0.0266	20	0.4013	72.8205
0.6185	0.0399	30	0.4096	79.7436
0.5898	0.0533	40	0.4439	124.3590
0.5579	0.0666	50	0.3908	71.4286

Framework versions

Transformers 4.47.1
Pytorch 2.5.1+cu121
Datasets 3.2.0
Tokenizers 0.21.0