whisper_large_v2_thi_dataset

This model is a fine-tuned version of openai/whisper-large-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3682
• Cer: 15.1611
• Wer: 26.8405

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: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer	Wer
0.7544	0.9998	3132	0.4112	17.4841	30.6117
0.4151	2.0	6265	0.3786	20.9872	35.5513
0.3044	2.9995	9396	0.3682	15.1611	26.8405

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2+cu118
• Datasets 2.19.0
• Tokenizers 0.19.1