whisper-base-ft-dialect-detection

This model is a fine-tuned version of openai/whisper-base on the common_language dataset. It achieves the following results on the evaluation set:

• Loss: 0.6761
• Accuracy: 0.6667

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: 32
• eval_batch_size: 16
• seed: 0
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	2	0.7001	0.3333
No log	2.0	4	0.6855	0.5556
No log	3.0	6	0.6761	0.6667
No log	4.0	8	0.6674	0.6667
No log	5.0	10	0.6609	0.6667
No log	6.0	12	0.6551	0.6667
No log	7.0	14	0.6507	0.6667
No log	8.0	16	0.6476	0.6667
No log	9.0	18	0.6457	0.6667
No log	10.0	20	0.6447	0.6667

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2+cu118
• Datasets 2.19.1
• Tokenizers 0.19.1