ashe194/700-fine-tuned-whisper-base-full

Whisper base fine tuned full - ashe194

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

• Loss: 0.0034
• Wer: 0.3422
• Cer: 0.3439
• Wer Ortho: 0.5124

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: 4e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer	Wer Ortho
No log	0.9935	76	0.0065	0.4277	0.2934	0.7385
No log	2.0	153	0.0035	0.3707	0.3717	0.5275
No log	2.9804	228	0.0034	0.3422	0.3439	0.5124

Framework versions

• Transformers 4.42.3
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1