dgx1_whisper_tiny_finetune_teacher_babble_noise_mozilla_40_epochs_batch_32

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

• Loss: 1.0819
• Wer: 38.0301

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: 0.0005
• train_batch_size: 32
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 1024
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_ratio: 0.2
• num_epochs: 40
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.622	7.35	500	0.8796	41.4282
0.151	14.7	1000	1.0639	41.4038
0.0146	22.06	1500	1.0756	38.3399
0.0004	29.41	2000	1.0791	38.0753
0.0003	36.76	2500	1.0819	38.0301

Framework versions

• Transformers 4.25.1
• Pytorch 1.12.1
• Datasets 2.8.0
• Tokenizers 0.13.2