PhoWhisper-small-vispeech-classifier-v3

This model is a fine-tuned version of vinai/PhoWhisper-small on the hr16/ViSpeech-Gender-Dialect-Classification dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.5291
• eval_accuracy: 0.8793
• eval_runtime: 180.5859
• eval_samples_per_second: 16.513
• eval_steps_per_second: 0.692
• epoch: 3.0
• step: 1470

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: 24
• eval_batch_size: 24
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.2
• num_epochs: 15
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1